CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
EBA/CP/2016/21
14 November 2016
Consultation paper
Guidelines on PD estimation, LGD estimation and the treatment of
defaulted exposures
1
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Contents
1. Executive Summary
3
2. Background and rationale
5
3. Draft Guidelines
32
4. Accompanying documents
100
4.1
Draft cost-benefit analysis / impact assessment
100
4.2
Overview of questions for consultation
115
2
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
1. Executive Summary
The draft guidelines presented in this consultation are one of the initiatives undertaken by the EBA to
reduce the unjustified variability and are part of a broader review of the IRB Approach that is carried
out by the EBA in accordance with the plan outlined in the Report on the regulatory review of the IRB
Approach published in February 2016 1. These guidelines are focused on the definitions and modelling
techniques used in the estimation of risk parameters for both non-defaulted and defaulted
exposures, whereas other regulatory products developed in the review process will clarify other
aspects related to the application of the IRB Approach. The EBA considers these clarifications and
harmonisation necessary in order to achieve comparability of risk parameters estimated on the basis
of internal models and to restore trust in these models by market participants while at the same time
preserving risk sensitivity of capital requirements.
The EBA has in its previous work identified a clear need for these guidelines, including in five reports
on the comparability and pro-cyclicality of capital requirements, developed in accordance with
Article 502 of Regulation (EU) No 575/2013 and published by the EBA in December 2013 2. These
reports confirmed significant discrepancies in risk parameters and own funds requirements across
institutions and jurisdictions that did not reflect differences in risk profiles but resulted from different
underlying definitions and certain modelling choices. These discrepancies were in part a
consequence of excessive flexibility incorporated in the IRB framework and are considered to be a
main driver in the loss of trust of internal models by observers, investors and other market
participants.
With regard to non-defaulted exposures the draft guidelines provide detailed clarifications on the
estimation of PD and LGD parameters. In the case of defaulted exposures institutions are required to
estimate LGD (so called LGD in-default) and expected loss best estimate (ELBE). As these parameters
are in fact part of LGD models the clarifications provided in the guidelines on the estimation of these
parameters are largely based on the requirements specified for the estimation of LGD for nondefaulted exposures. In addition, the guidelines specify aspects common for all risk parameters such
as the use of human judgement both in the development and in the application of the internal
models, appropriate margin of conservatism that should be incorporated in risk parameters as well
as regular reviews of the models in order to ensure timely implementation of necessary changes in
case of deteriorated performance of the models. The aim of the guidelines is therefore to harmonise
the concepts and methods used today.
As it is expected that these guidelines may lead to material changes in numerous rating systems used
currently by institutions sufficient time has to be granted for their implementation. The proposed
deadline for implementation is end-2020 as already specified in the Opinion on the implementation
1
https://www.eba.europa.eu/-/eba-sets-out-roadmap-for-the-implementation-of-the-regulatory-review-of-internalmodels
2
http://www.eba.europa.eu/-/eba-publishes-reports-on-comparability-of-risk-weighted-assets-rwas-and-pro-cyclicality
3
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
of the regulatory review of the IRB Approach published by the EBA in February 2016 3. This opinion
describes the envisaged phasing-in approach and the specified deadline refers to implementation of
all changes stemming from the regulatory review of the IRB Approach.
Next steps
The draft guidelines are published for the 3 months consultation period. At the same time the EBA is
planning to carry out a qualitative survey across institutions in order to assess the impact of the
proposed requirements on the rating systems. The responses received during the consultation period
and the results of the survey will be taken into account when specifying the final guidelines.
In addition, the EBA is planning to consult on the draft RTS on the nature, severity and duration of
economic downturn developed in accordance with Article 181(3)(a) of Regulation (EU) 575/2013. As
these RTS will be closely related to the estimation of downturn LGD some additional changes may be
introduced in the final guidelines on the basis of the feedback received during these consultations.
3
https://www.eba.europa.eu/-/eba-sets-out-roadmap-for-the-implementation-of-the-regulatory-review-of-internalmodels
4
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
2. Background and rationale
Introduction
The concept of the Internal Ratings Based Approach (IRB Approach) for credit risk was first
introduced by the Directive 2006/48/EC of 14 June 2006 (part of what was known as the Capital
Requirements Directive – CRD), later replaced by Regulation (EU) No 575/2013 (Capital Requirements
Regulation – CRR). The CRR introduced a number of mandates for the EBA to develop technical
standards and guidelines to supplement the primary legislation in order to ensure more harmonised
application of the IRB requirements.
In this regard and in accordance with Article 502 of the CRR, EBA published in December 2013 a set
of five reports on the comparability and pro-cyclicality of capital requirements, presenting the results
of a study that EBA conducted on the comparability of the risk estimates and capital requirements
including the analysis of the factors that contribute to the discrepancies among institutions. Based on
the results, the EBA concluded that further guidance was needed, as current practices differed
significantly across countries and institutions. Consequently, the EBA initiated work to provide
further regulatory guidance and these draft Guidelines on PD estimation, LGD estimation and the
treatment of defaulted assets (GL) are one of the initiatives and specifically target the identified
significant discrepancies in the methodologies underlying risk estimates.
The sources of discrepancies identified in the area of modelling were mostly related to different
definitions of the main concepts underlying the risk parameters and different modelling choices that
were possible due to a large degree of flexibility incorporated in the IRB framework. In addition,
different understanding of regulatory requirements was also observed.
These draft GL are therefore focused on aligning the terminology and definitions, in particular in
relation to metrics such as default rate or realized LGD that are the basis for estimation of risk
parameters. Furthermore, the draft GL provide clarification on the application of certain regulatory
requirements that were until now interpreted in various ways and specify principles for the
estimation of risk parameters, including those applicable to defaulted exposures. Although the draft
GL may limit certain modeling choices they are focused on the elements that lead to non-risk based
variability and intend to preserve sufficient flexibility to ensure risk sensitivity of the models.
Therefore, the draft GL do not prescribe any specific estimation methodology where different
approaches may be appropriate for different portfolios in order to reflect different risk profiles.
The main objective of the draft GL is to provide the rules that will lead to comparability of the model
outcomes. Differences in risk parameters between institutions should ideally reflect differences in
the underlying risk rather than different modelling choices. In addition, clearer rules in that regard
will limit the possibility for regulatory arbitrage. Other aspects of the models that are not explicitly
prescribed in the draft GL, such as the choice of risk drivers and estimation methodology, will have to
be justified on the basis of the risk profile of the portfolio covered by the model as well as the credit
and recovery policies and efficiency of these processes.
5
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
As the draft GL are part of a broader review of the IRB Approach carried out by the EBA they do not
address all identified sources of RWA variability. The GL focus on aspects related to modelling of
parameters such as PD, LGD, ELBE and LGD in-default whereas other elements, including the
definition of default that these parameters should be based on, rating processes, data quality
processes and other aspects of the application of the IRB Approach are addressed in other regulatory
products. The EBA’s plan for the regulatory review of the IRB Approach has been outlined in the
report published in February 2016 4. The planned regulatory products will affect nearly all aspects of
the IRB Approach and it is expected that as a consequence they will be able to significantly reduce
the unjustified RWA variability which is deemed to stem from the lack of sufficiently specified
requirements with regard to certain aspects of the IRB Approach, where such specification is deemed
necessary in order to achieve the objective of a Single Rulebook, as well as to regain public trust in
the use of internal models.
It has to be stressed that these GL include numerous references to the RTS on IRB assessment
methodology which set conditions for competent authorities to assess the rating systems of the
institutions. Although the RTS is addressed to competent authorities it is binding also to institutions.
Therefore, these GL and the mentioned RTS should be read together as many aspects related to
modelling have already been clarified in the RTS and in these cases the provisions are not repeated in
the GL. When implementing any changes in the rating systems stemming from the regulatory review
of the IRB Approach, and also subsequently on a continuous basis, institutions should take into
account not only these GL but also provisions included in other related regulatory products, in
particular the RTS on IRB assessment methodology, the RTS on materiality threshold for past due
exposures, the GL on the application of the definition of default and the RTS on the nature, severity
and duration of economic downturn. The consultation paper on the latter RTS is planned to be
published later this year. It will consult on the specification of the downturn period as well as other
aspects related to the estimation of downturn LGD.
Neither these GL nor any other of the EBA’s regulatory products address the issue of the scope of
application of the IRB Approach and modellability of low default portfolios. These aspects are
currently under consideration at the international level by the Basel Committee on Banking
Supervision (BCBS) and may subsequently be incorporated in the European legal framework by
relevant changes to the CRR. Regardless of these possible developments the provisions included in
these GL and other related regulatory products will apply to these models and portfolios that remain
within the scope of the IRB Approach.
This Consultation Paper seeks feedback from the industry on the proposed guidance on PD
estimation, LGD estimation and the treatment of defaulted assets with respect to the suitability for
lowering unjustified variability of risk parameters, but also with respect to operational issues which
might arise when following the proposed GL. A more detailed rationale for the proposed provisions is
presented in the subsequent sections.
4
https://www.eba.europa.eu/documents/10180/1360107/EBA+Report+on+the+regulatory+review+of+the+IRB+Approach.p
df
6
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Chapter 4: General estimation requirements
This section covers policy proposals for segmentation principles, data requirements, human
judgement in model development and margin of conservatism.
Segmentation principles
The segmentation principles aim to provide guidance on the highest level of rating system design.
These principles are particularly relevant in the case of changes in the scope of application of certain
rating systems, for example where an existing rating system is rolled out to an acquired portfolio or a
portfolio that is otherwise not yet treated under the IRB Approach. The draft GL require in this regard
among others the fundamentally comparable availability of credit related information, meaning that
with respect to the obligors or exposures to which the rating system is extended the relevant
information has a similar nature and is available for the purpose of rating assignment or at least is
possible to obtain. For instance, as the information available for business clients and for natural
persons is fundamentally different, these should not be covered by the same rating system.
Data requirements
Good quality of data is a fundamental condition for developing a robust rating system. The data
requirements in this general part apply to model development and application as well as risk
parameter quantification for all risk parameters and contain clarifications regarding the assessment
of accuracy, completeness and appropriateness of data. Data requirements specific to PD or LGD
estimation (or LGD-in-default or ELBE estimation) are described within the according sub-chapters of
these draft GL.
Human judgement in model development
Development of a robust rating system cannot be a purely statistical process, but to some extent has
to involve also human judgement in order to make sure that the models are appropriate for current
and foreseeable portfolios and conditions and that the models are acceptable for business users.
Expert judgement may be necessary in particular with respect to the verification of model
assumptions and whether these are in line with economic expectations, the design of the model, the
choice of risk drivers, etc. However, in order to ensure a high quality of the models the expert
judgement has to be appropriately documented and justified. This way the judgmental elements of
the model can be appropriately challenged and verified both by the validation function as well as by
competent authorities. Therefore, this part of the draft GL clarifies the requirements regarding
human judgment in model development and with respect to the documentation of human judgment
in model development.
Margin of conservatism (‘MoC’)
The draft GL include principles to be followed by institutions in the identification, quantification,
reporting and documentation of an adequate margin of conservatism. As a general concept
institutions are required to address the identified deficiencies in data or methods via appropriate
adjustments and margin of conservatism. An appropriate adjustment consists in rectifying the
7
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
identified errors, for instance missing data points are filled in with the most probable information or
the inaccuracies in data are corrected. The objective of the appropriate adjustment is to achieve the
possibly most accurate estimates. However, as the appropriate adjustment is an estimate due to data
deficiency, additional MoC has to be added to address the uncertainty related to this estimation.
Moreover, MoC aims at addressing all errors that cannot be rectified through appropriate
adjustment and any other uncertainties related to the estimation of risk parameters.
Example for appropriate adjustment and margin of conservatism
Assume a change in regulatory requirements regarding the materiality threshold for detecting
defaults triggered by the 90 days past due (‘dpd’)-criterion. Assume the considered institution has
stored the information regarding outstanding exposure only monthly and not daily. Thus the date of
default as well as the EAD cannot be recollected historically. Even though some calculations could be
made based on the monthly data, the 90 days could have been reached during a month and the
default could not be visible based on monthly data due to repayments. Therefore the institution
decided to set up a parallel default detection according to the new trigger and estimate MoC starting
from the difference of the amount of defaults detected according to the old trigger and the new
trigger.
An appropriate adjustment could be derived as follows: Calculating the relative change in the
number of defaults triggered according to the old 90dpd and the new 90dpd criterion compared to
all defaults (i.e. taking into account all triggers) on a monthly basis. Thus an average correction factor
can be estimated and applied backwards. As a result the number of defaults according to the new
90dpd trigger can be estimated for the available historical data.
The additional MoC could be derived for example from the 90% confidence interval around the
average of the new default rates. A non-exhaustive and exemplary list of triggers for appropriate
adjustments and MoC can be found in Annex I. However, the methods described in this annex are
only examples, institutions may use other methods for deriving MoC if these are deemed more
appropriate.
Chapter 5: PD estimation
The guidelines on PD estimation aim to provide among others more detailed guidance on the
calculation of observed default rates and on the estimation of the long run average default rate.
Moreover, it clarifies how risk drivers and rating criteria should be chosen and which requirements
should be fulfilled in case ratings serve as input to the PD estimation. Other aspects touched upon
are the rating philosophy and how the long run average default rates relate to the final PD estimation
of a grade or pool.
General requirements specific to PD estimation
One of the main aspects addressed in the general part of the chapter on PD estimation is the
requirement that each natural or legal person that has exposures within the scope of the IRB
Approach should be rated including where there is unfunded credit protection. The rational for this is
that for the purpose of model development, risk quantification and validation the assignment to a
8
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
pool or grade after the application of unfunded credit protection would bias risk differentiation and
default rate calculation.
The other main policy proposal provides clarification on the updates of rating relevant information.
In a situation where an institution receives relevant new information this update should be
incorporated into the rating assignment of the according obligor as soon as possible. However, in
case of specific data e.g. relevant balance sheet data, new information will be available for a major
share of the obligors at the same time of the year due to common reporting dates and updating the
relevant information and reviewing the according rating assignments for a major share of the
portfolio requires a realistic time window. Therefore, it is required that a resulting review of the
rating assignment should not be made later than 3 months after the new information becomes
available. Moreover, for retail rating systems where information becomes available in other IT
systems than the rating relevant IT systems, this information should be taken into account in the next
rating review.
Data requirements specific to PD estimation
The data requirements specific to PD estimation are split in two sub-sections, namely data
requirements for the purpose of default rate calculation and data requirements for the reference
data set for model development. This split is motivated by the according structure of the CRR, where
data requirements for risk quantification are treated in sub-section 2 and data requirements
regarding models are treated in sub-section 1 of section 6 in Chapter 3 of Part Three, Title II of the
CRR. In particular, institutions may, subject to certain requirements, use a definition of default other
than the one specified in Article 178 of the CRR. The main difference between the data requirements
for default rate calculation and those for model development is that non-comparability for the
purpose of risk quantification should not lead to data exclusion, but trigger an adjustment. However,
for the purpose of model development material observed differences in the key characteristics
should be avoided, for example by using another sample.
Data requirements for default rate calculation
For the default rate calculation, the draft GL clarify among others that all data relevant for identifying
the non-defaulted exposures at the beginning of a one-year observation period has to be available as
well as all relevant default information as required in Articles 178 of the CRR. This clarification was
considered necessary as the emphasis is often put on the data collection related to the defaulted
exposure but for the purpose of default rate calculation the accurate and complete identification and
data collection for non-defaulted exposures is equally important.
Data requirements for the reference data set for the purpose of model development
One of the aspects related to data used for model development that should be properly considered
in the model development is the choice of reference points in time where risk drivers and rating
criteria should be evaluated. The draft GL require the use of appropriate points in time, which may
be different for different risk drivers.
9
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Observed default rates
The section on the calculation of observed default rates clarifies in more detail which obligors should
be taken into account in the nominator and denominator for the purpose of calculating a one-year
default rate and requires certain analysis depending on the method an institution chooses for
averaging over a series of one-year default rates.
Calculation of the one-year default rate
While the general calculation of a one-year default rate is already outlined in the CRR and the RTS on
IRB assessment methodology, the draft GL clarify this calculation for a number of specific situations.
In particular, the denominator should contain the obligors of the considered model or calibration
segment with any credit obligation at the beginning of the observation period. Where obligors whose
obligations stem solely from non-credit products fall under the scope of application of the considered
model and are treated in accordance with the institution’s internal default definition, then these
should form a separate pool in the rating system not to bias the default rate of obligors with credit
facilities. Similarly, with regard to obligors or facilities with just committed but undrawn credit lines
these might have to be treated in a separate pool in the rating system to avoid lowering unduly the
default rate of drawn credit lines. It has to be noted as well that an obligor can also default if there is
no repayment obligation during the observation period, as for example a bankruptcy notice might
occur at any time.
Calculation of the observed average default rate
The draft GL require institutions to justify their approach to calculating the average of one year
default rates taking into account, in particular, analysis on the effect of short term contracts or
specific reporting dates.
Long-run average default rate
Regarding the long run average default rate, the GL clarify that this should be calculated as the
average of observed one year default rates if the historical observation period is representative of
the likely range of variability of one-year default rates and, in particular, if the historical observation
period contains a downturn period. If the one year default rates are not representative of the likely
range of variability, in particular, if no downturn period is contained, then institutions should
estimate the long run average default rate by estimating an appropriate adjustment to the average
of observed one year default rates.
In order to limit possible variability stemming from the application of this concept a benchmark is
proposed, namely the maximum of the average of one year default rates over the most recent five
years and the average of one year default rates over the whole available observation period.
Institutions may still estimate long-run average default rates below this benchmark, but this should
be duly justified and eventually trigger additional margin of conservatism.
10
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
PD estimation methodologies
The last section in this chapter is concerned with PD estimation methodologies and contains in
particular policies for the use of third-party ratings in PD estimation, the design of grades and pools
and for the process of assigning PD estimates to grades and pools. One of the issues that were
clarified is that the calibration sample, namely the sample which is used for the process of assigning
long run average PD to grades or pools, should be comparable to the current portfolio in terms of
obligor and transaction characteristics but should reflect at the same time the likely range of
variability.
Chapter 6: LGD estimation
General requirements for LGD estimation
General requirements for LGD estimation outline the scope of methodologies that can be used for
the purpose of LGD estimation. In this context workout LGD is considered to be the main, superior
methodology that should be used by institutions. It is essential that LGD estimates are based on the
institutions’ own loss and recovery experience in order to make sure that the estimates are adequate
for the institutions portfolios and policies and in particular that they are consistent with the recovery
processes. Therefore, although internal experience may be supplemented with external data,
institutions should not use methodologies that are based only on external data, such as so called
market LGD and market implied LGD which are based on the market prices of financial instruments
such as marketable loans, bonds or credit default instruments.
An alternative methodology that is available for retail exposures and purchased corporate
receivables is deriving LGD estimates from realised losses and appropriate estimates of PD. However,
also in this case institutions should ensure that the estimates are sufficiently robust. This is ensured
where both realised losses and PD estimates meet all relevant requirements. In particular, in order to
ensure comparability of LGD estimates based on this methodology with other LGD estimates the
calculation of total losses has to be consistent with the concept of economic loss used for the
purpose of workout LGD.
Data requirements for LGD estimation
Reference Data Set
As the LGD estimates should be based on the institution’s own experience it is important that all
relevant data is properly recorded and stored. The scope of data necessary for proper LGD
estimation is very broad and entails not only the date of default and all cash flows and events after
default but also all relevant information about the obligors and transactions that could be used as
risk drivers in the model development.
One of the most important risk drivers in the LGD estimation is the existence of collaterals. As the
observed and estimated recovery rates relate to the value of the collateral, the timing and type of
valuation is a key aspect that may significantly influence the estimates. In order to ensure that the
estimates are adequate for the existing portfolio it is essential that the information about the
11
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
valuation is used consistently in the LGD estimation and in the application of LGD estimates. Hence,
as the LGD estimates are applied to non-defaulted exposures, institutions should use the valuation
from before the moment of default in the estimation of LGD. Although a more up-to-date valuation
may be available that was performed after default or in relation to default this information should
not be used in the estimation of LGD because this valuation is not comparable to the valuation of
collateral for non-defaulted exposures. Valuations performed after default are often more
conservative and hence the use of such information could lead to overestimation of recovery rates.
Representativeness of data
Representativeness of data may influence the accuracy of the estimates; where the underlying
historical data is less representative to the current portfolio the estimates may be less adequate. The
dimensions of representativeness include the scope of application of the model, distribution of the
relevant risk drivers, definition of default and lending standards and recovery policies. However, even
where historical observations are not fully representative they still contain valuable information.
Therefore, non-representativeness should lead to appropriate adjustments, where possible, and
additional MoC but should not be a justification for excluding the data from the estimation process.
This is also consistent with Article 181(1)(a) of the CRR which requires the use of all observed
defaults.
Furthermore, in accordance with Article 179(1)(d) of the CRR the economic or market conditions that
underlie the data should be relevant to current and foreseeable conditions. In this context it is
considered that all market and economic conditions experienced in the past are within the scope of
foreseeable conditions. As the purpose of the own funds requirements is to address the unexpected
loss even if extreme events were observed in the past these should not be excluded from the
estimation sample.
Calculation of economic loss and realised LGD
Definition of economic loss and realised LGD
The concepts of economic loss and realised LGD are the basis for LGD estimation and any differences
in the calculation may lead to significantly different and non-comparable LGD estimates. Therefore, it
was considered essential to specify these concepts in detail, including the treatment of unpaid late
fees, interest and additional drawings after default, discounting rate and costs. The specification was
based on the definitions included in the CRR but provides more detailed clarifications on the practical
application of these definitions.
Treatment of unpaid late fees, interest and additional drawings after default
With regard to the treatment of unpaid late fees, interest and additional drawings after default it was
considered that in order to reflect correctly the level of loss these should also be included in the
calculation. Otherwise, if only recoveries related to these events were included but the measure of
loss was not increased by increased credit obligation this could lead to underestimation of LGD and in
some cases to negative realised LGDs. In this sense the treatment of fees and interests should be
12
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
consistent as the economic sense of these measures is similar and it may only depend on the
institution’s pricing policy whether the obligors in default are charged with interests or with fees.
Furthermore, in specifying the treatment of additional drawings, fees or interests after the moment
of default it was considered that the resulting measure of realised LGD should be consistent with the
exposure value that will be used for the purpose of calculation of capital requirements. Hence, where
additional drawings after default are included in the conversion factors the outstanding amount in
the denominator of the realised LGD should also include such drawings. However, in the case of retail
exposures, where the conversion factors do not include any drawings after the moment of default
the denominator of the realised LGD should also reflect only the outstanding amount at the moment
of default. As a result, lower exposure value (based on lower conversion factors) will be
compensated by higher LGD.
Similarly, in the case of additional fees or interests that are capitalised after the moment of default,
as these are generally not included in the conversion factors they should not increase the
outstanding amount at the moment of default in the denominator of realised LGD. It is proposed that
such fees and interests should be treated similarly as costs, i.e. they should increase the measure of
exposure and loss in the numerator of the realised LGD, just as all recoveries related to those fees or
interests, adequately discounted, should be included in the calculation of economic loss. At the same
time, where these additional fees cover costs that were incurred by the institution, these should not
be calculated twice.
In order to keep consistency between the LGD and exposure value including conversion factors
different calculation of the denominator of the realised LGD is proposed depending on whether
additional drawings after default are included in the estimation of conversion factors or not.
However, this optionality does not refer to the measure of economic loss, i.e. the numerator of the
realised LGD, as this should be an objective value that adequately reflects the actual value of loss
experienced by the institution.
Discounting rate
The EBA has considered various possibilities with regard to the discounting rate and analysed various
practices in that regard. Differences in approaches used by institutions range from the use of
discounting factors based on effective interest rates on the underlying loans, different add-ons in the
range of 0 to 10% and even higher in some cases on top of different underlying internal and external
interest rate benchmarks. As different approaches are currently adopted by institutions the
discounting factor was recognised as one of the main drivers of non-risk based variability of the LGD
estimates. The proposed solution of using interbank funding rates and a 5% add-on has the
advantage of being simple and contributing to increased comparability of LGD estimates. It is
considered appropriate that the discounting rate should not depend on the credit standing of the
institution and hence the discounting rate does not reflect funding costs but is rather focused on the
uncertainty inherent in the recovery processes and the time value of money.
13
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Direct and indirect costs
With regard to costs to be included in the calculation of the economic loss the proposal follows the
distinction into direct and indirect costs depending on whether they are directly attributable to a
given exposure. It has been specified that all direct costs should be considered material whereas
immaterial indirect costs may be excluded from the calculation.
In order to reflect the full level of loss it is proposed that institutions should look into costs not only
after the moment of default but also before that date. If the costs incurred by the institution due to
diminished credit quality of the exposures but before recognition of default are not included in the
exposure value at the moment of default they should be taken into account in the calculation of
economic loss. Otherwise these costs would not be accounted for in the estimation of risk
parameters and LGD would be underestimated.
Long-run average LGD
Historical observation period
The specification of the historical observation period is based on the assumption that it should be as
broad as possible and should contain data from various periods with differing economic
circumstances. These differing economic circumstances refer not only to the moment of default but
also to the moment of realising recoveries from different sources. In this context it was deemed
inappropriate to allow elimination of any data that reflects an institution’s internal experience as this
would lead to a loss of valuable information. Hence, it was specified that all available internal data
should be taken into account in the long-run average LGD. This internal experience may be
additionally supplemented by external data where necessary.
Calculation of long-run average LGD
It has been clarified that the long run average should be calculated as an arithmetic average of
realised LGDs on all observations from the specified historical observation period and that it should
be weighted by the number of defaults. The only exemption from this rule under the CRR is specified
in Article 181(2) of the CRR according to which institutions may use higher weights to more recent
data in the case of retail exposures. However, the use of this exemption requires appropriate
justification and evidence that it leads to better LGD estimates.
Treatment of incomplete recovery processes
In accordance with Article 181(1)(a) of the CRR all observed defaults have to be taken into account in
the calculation of long-run average LGD; hence, also incomplete recovery processes should be
included. These incomplete processes carry valuable information in particular about the most recent
observations and the cases that are particularly difficult and therefore require longer recovery
processes. Exclusion of this information would not only lead to loss of relevant, up-to-date
information but could also lead to underestimation of LGD and therefore this was considered
inappropriate.
14
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
However, in order to obtain a realistic value of long-run average LGD the incomplete recovery
processes should be included with future recoveries that are expected to be realised. The value of
future recoveries is not and an objective, observed measure but has to be estimated based on the
recoveries factually observed on those cases that are already closed. As a result, the ‘long-run
average LGD’ will also be a measure that is not fully objective as it contains components that are
estimated.
In order to obtain a fully objective measure it is proposed that institutions should also calculate the
‘observed average LGD’ taking into account realised LGDs only on those defaults that are related to
closed recovery processes and those that reached a certain threshold in terms of the time in default,
i.e. maximum length of the recovery process during which additional recoveries can be reasonably
expected. Although this objective measure will not include any elements of estimation it has to be
kept in mind that it may not reflect the real experience correctly as the cure and high-recovery cases
may be overrepresented. More difficult cases usually stay longer in recovery processes therefore
they will more likely not be included in the ‘observed average LGD’.
Therefore, the ‘observed average LGD’ has to be adjusted to account for the most recent experience
based on the incomplete recovery processes. For this purpose institutions should estimate the most
likely future recoveries on cases where the processes are not yet complete. As such estimates can
only be provided where sufficient data exists to support them it is proposed that institutions should
only estimate future recoveries until a certain point in time, i.e. maximum length of the recovery
process during which recoveries are actually observed on similar cases.
Treatment of cases with no loss or positive outcome
Finally, it is also proposed that where the calculation of realised LGD results in a negative number, i.e.
where profit has been realised on a defaulted exposure, this should be floored at zero. This floor
should be applied at the level of individual observation as it would not be appropriate to allow any
netting effects in that regard.
LGD estimation methodologies and risk drivers
The GL do not prescribe any specific methodology that should be used in the estimation or LGD. It
has been recognised that various methodologies may be valid, depending on specific circumstances,
portfolios and processes. However, it was considered appropriate to specify certain principles that
should be adhered to regardless of the methodology that is chosen.
As part of these general principles it has been specified which types of potential risk drivers should
be taken into account by institutions. These include factors related to transactions and obligors but
also to institutions themselves, in particular in terms of the organisation of the recovery processes as
well as external factors such as legal frameworks, especially where models apply to exposures in
various countries. It is important that institutions duly analyse potential risk drivers and choose those
that meaningfully differentiate risk of transactions. In addition, the risk drivers should be analysed at
an appropriate reference date that is representative of the realisations of the given risk driver within
a year before default. This has the purpose of ensuring consistency between the estimation and the
application of LGD, where the estimates will apply to non-defaulted exposures.
15
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Treatment of collaterals in LGD estimation
Eligibility of collaterals
Based on Article 108(2) of the CRR it has been clarified that for the purpose of LGD estimation
institutions may take into account any type of collateral as long as requirement of Article 181(1)(f) of
the CRR is met. It has been further clarified in the RTS on IRB assessment methodology that in order
to meet this requirement the institution’s internal policies should be at least fully consistent with the
requirements of Section 3 of Chapter 4 of the CRR with regard to legal certainty and regular valuation
of collateral. It is also envisaged that institutions may use for the purpose of LGD estimation specific
types of collaterals that are not explicitly described in Chapter 4 of the CRR. In these cases the
policies and procedures relating to internal requirements for valuation and legal certainty should be
appropriate to the respective type of collateral.
Inclusion of collaterals in the LGD estimation
The existence of collateral is one of the main aspects that affects the recovery processes and their
results. As in accordance with Article 179(1)(a) of the CRR institutions are required to incorporate in
their LGD estimates all relevant data, information and methods it has been specified that information
on at least the main types of collaterals used for a given type of exposures should be considered
relevant and included in the LGD estimates. This means, however, that for the main types of
collaterals the requirement of Article 181(1)(f) of the CRR will have to be met as specified above.
Furthermore, the draft GL specify general principles for reflecting the effect of collaterals in the LGD
estimates without prescribing any specific methodology. These principles include avoiding bias in the
LGD estimates that may stem from inappropriate treatment of cash flows realised with the use of
collaterals as well as from inappropriate valuation of the collateral.
Cash flows from collaterals
It has been clarified that for the purpose of calculation or realised LGD and LGD estimation the
recoveries realised with the use of collaterals have to be treated as such regardless of the form of the
realisation of the collateral. This could include not only workout processes through court proceedings
but also sale of the collateral by the obligor himself, normally with the consent of the institution, or
the sale of the credit obligation where the collateral is reflected in the price.
Broader clarification has been provided for a specific case where the collateral is repossessed by the
institution in exchange for decreasing the credit obligation. It is proposed that this event should be
treated as recovery, as from the economic perspective such event is equivalent to receiving a cash
payment and investing it in an asset. However, as the value of repossession does not always reflect
accurately the market value of the asset, this uncertainty should be addressed by applying an
appropriate haircut to the value of repossession. Although institutions may have different strategies
with regard to the repossessed assets, and in particular in some cases they may decide to keep the
asset on their balance sheet for speculative purposes, these different strategies should not influence
the value of the recovery. Therefore it has been specified that the haircut should be estimated with
the assumption that that the institution intends to sell the repossessed asset as soon as it is
16
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
reasonably possible. Wherever sufficient past experience with regard to repossession of collaterals
exists the haircuts should be supported by historical observations and regularly backtested. In the
absence of such experience the assessment will have to be performed on a case-by-case basis, but
this will require more conservatism as such assessment will be less reliable.
Downturn adjustment
Clarification of the determination of downturn period based on its nature, severity and duration will
be provided by the RTS developed on the basis of Article 181(3)(a) of the CRR. The currently
proposed text of the draft GL refers to these RTS but does not provide any further clarification on
how to calculate the downturn adjustment to LGD estimates. It is assumed that various
methodologies may be relevant in different situations. It is however still under consideration
whether more guidance in that regard should be provided in the draft GL.
Chapter 7: Estimation of risk parameters for defaulted exposures
General requirements specific to Elbe and LGD in-default
The treatment of defaulted assets was identified as one of the major drivers of variability of the own
funds requirements across institutions. Clarification has already been already provided in the RTS on
IRB assessment methodology, in particular Article 54(2)(c), that the direct estimation of LGD indefault should be consistent with the LGD for non-defaulted exposures in order to avoid potential
cliff effects. Following this approach it has been further clarified in the draft GL that for the purpose
of estimating ELBE and LGD in-default institutions should use the same estimation methods used for
estimating LGD on non-defaulted exposures as they are in fact part of the LGD model. Thus, Chapter
7 generally refers to the requirements on LGD estimation set out in Chapter 6 as well as general
estimation requirements set out in Chapter 4 and the requirements on the application of risk
parameters specified in Chapter 8 and provides guidance only on those specific aspects where
different treatment for defaulted assets loss rate estimation is justified.
Data requirements specific to LGD in-default and Elbe
As for non-defaulted exposures, ELBE and LGD in-default estimates should be based on the
institutions’ own experience. The scope of data necessary for proper ELBE and LGD in-default
estimation not only includes those required for LGD for non-defaulted exposures but also all relevant
operational information obtained during the recovery process and, in particular, at each reference
date used in the estimation. This implies that for the purpose of the treatment of defaulted assets
institutions should additionally store relevant risk drivers, including those that become relevant after
default, and outstanding exposure amounts at each reference date.
Reference dates
The difference between the LGD in-default and the ELBE is used for computing the risk weight
according to Article 153(1)(ii) of the CRR which is then applied to the current outstanding exposure
amount in order to obtain the risk weighted exposure amount. Moreover, the ELBE is compared to
credit risk adjustments for IRB shortfall / excess purposes where credit risk adjustments are again
17
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
computed with respect to the current value of exposures. Thus, for the purpose of computing
realised LGDs for defaulted exposures institutions should use reference points in time that will be
relevant for the current outstanding obligations of defaulted exposures.
The concept of current outstanding exposure is clearly defined in Article 166(1) of the CRR and
should be used also for defaulted exposures in the application of the ELBE and LGD in-default.
However, given data limitations, the continuous concept of current exposure amount may not be
suited for estimation purposes. The draft GL therefore suggest that institutions should set discrete
relevant reference dates at which the realised LGDs should be computed. This way it should be
feasible to estimate the parameters for defaulted exposures that are appropriate for their current
status. In order to ensure adequacy of the estimates institutions should set the reference dates
according to the recovery pattern observed on a specific type of exposures , where such reference
dates may be either event based, e.g. linked with the realisation of collateral, or may reflect certain
time periods during which exposures have been in-default.
For the purpose of application of the estimated LGD in-default and ELBE to a given defaulted exposure
in the current portfolio institutions should first evaluate which reference date is relevant for the
exposure under consideration. The risk parameters to be assigned to the defaulted exposure under
consideration should be calculated as the product of the ELBE (or the LGD in-default) relevant at the
selected reference date in percentage terms and the current outstanding exposure amount.
Calculation of realised LGD and long-run average LGD for defaulted exposures.
One major difference between the calculation of realised LGDs on defaulted and non-defaulted
exposures is that the former should be performed at each relevant reference date rather than at the
date of default, as described above. Other than this, in order to calculate realised LGDs for defaulted
exposures institutions should follow the same requirements as those set in Chapter 6. This implies
that institutions should calculate for each defaulted exposure in the RDS the realised LGDs according
to each reference dates relevant for estimation purposes.
Another important aspect clarified in this section concerns the treatment of incomplete recovery
processes for the purposes of calculation of long-run average LGD for defaulted exposure. The
approach to the long-run average LGD is aligned to the one prescribed in section 6.4 with the
exception that incomplete recovery processes could be used only for those reference dates beyond
which factual recoveries and costs are observed. The rationale for this exception is that to avoid a
circular reference in the estimates. In fact, incomplete recoveries processes on which we are
estimating ELBE and LGD in-default should not take part in the estimation itself and will be therefore
excluded if only those incomplete recovery processes are taken into account for which later
reference dates are relevant.
Risk drivers
This section specifies which types of potential risk drivers should be taken into account in estimating
ELBE and LGD in-default on top of those used for non-defaulted exposures. Article 54(2) of RTS on IRB
assessment methodology prescribes that the LGD in-default and ELBE estimation methods take into
account the information on the time in-default and recoveries realised so far. In this respect, the
18
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
draft GL clarify that the information on recoveries realised so far and on time in-default may be taken
into account either directly, as risk drivers, or indirectly, in setting the reference dates for estimation
purposes. Moreover, in order to ensure that the information after default is timely and efficiently
taken into account, it is clarified that the relevance of risk drivers should be re-evaluated for the
relevant periods after default until the date of termination of the recovery process. This implies, for
example, that new risk drivers might become relevant after the date of default.
Specific requirements for ELBE estimation
In accordance with Article 181(1)(h) of the CRR the ELBE estimation methods should take into
account all currently available and relevant information and, in particular, consider current economic
circumstances and exposure status. Taking this into consideration the draft GL clarify that the ELBE
should not include any MoC as this would not be in line with the best estimate concept. Adding
conservativeness, in fact, does not increase the accuracy of the estimates but rather covers for the
risk that the estimates might be too optimistic.
Current economic circumstances
For the purposes of considering current economic circumstances the draft GL clarify that institutions
should take into account economic factors, including macroeconomic and credit factors, which are
relevant for the type of exposures under consideration. In this context, it is specified in the GL that
institutions should obtain their ELBE estimates through an adjustment to the long run average LGD for
defaulted exposures, consistently with the estimation of LGD in-default, such that a meaningful
application of the risk weight formula is ensured. The difference between LGD in-default and ELBE
determines the level of unexpected loss as it is used for computing the risk weight. Therefore, the
draft GL aim to ensure consistency of the estimation approaches used for the two risk parameters.
The difference between the ELBE and LGD in-default should reflect different economic conditions
considered, current economic circumstances for ELBE versus downturn conditions for LGD in-default,
and the application of the MoC for LGD in-default.
Concerning the calibration of the adjustment to current economic conditions examples are provided
of the approaches that could be used. This entails using risk drivers in the model that are sensitive to
economic factors, including in such a way current economic circumstances in the application of the
ELBE estimates, or including the economic factors directly in the model. Neither of these approaches
is prescribed as different approaches may be appropriate to different circumstances. However,
regardless of the approach used for the calibration of the adjustment to current economic
circumstances institutions should document the split of their ELBE estimates into the long run average
and the adjustment to current economic circumstances. This adjustment should be consistently
applied across portfolios and over time and institutions should document its rationale and
calculation.
Relations of ELBE with specific credit risk adjustments
The calculation of the IRB excess/shortfall in accordance with Article 159 of the CRR is based on a
comparison between the expected losses and credit risk adjustments. In this context using provisions
as ELBE estimates is very frequent practice observed within European institutions. However, this
19
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
approach is considered inappropriate as it does not ensure compliance with the CRR requirements. In
order to ensure consistency between the ELBE and LGD in-default estimates, the draft GL constrain
the use of provisions as ELBE to two specific circumstances. The first refers to those cases where a
provisions model respects all the requirements for own LGD estimates set in the CRR and in these
draft GL or when they can be adjusted to meet those requirements, in particular those related to the
concept of economic loss. The second possibility refers to those cases where provisions are
individually assessed, and so there is no model behind them. In these circumstances the information
of the individual assessment could be used as a potential reason for an override of the ELBE model
outcomes where institutions are able to prove that this improves the accuracy of the estimation. For
this purpose, individually assessed provisions should be adjusted in such a way to be consistent with
the requirements on economic loss set in these draft GL.
Specific requirements for LGD in-default estimation.
According to Article 54(1) of the RTS on IRB assessment methodology institutions may estimate LGD
in-default either directly or as a sum of ELBE and an add-on that captures the unexpected loss that
might occur during the recovery period. Irrespective of the approach it is expected that the method
for the estimation of LGD for exposures in default should consider a possible adverse change in
economic conditions during the expected length of the recovery process according to Article 54(2)(a)
of the RTS on IRB assessment methodology. It is clarified in the draft GLs that in order to reflect the
adverse change in economic conditions institutions should reflect in their LGD in-default estimates at
least downturn conditions. This is in line with Article 181(1)(b) of the CRR as LGD in-default is in fact
part of an LGD model. However, LGD in-default reflecting downturn conditions does not preclude the
inclusion of additional sources of uncertainty that are not related to economic conditions during the
recovery process. For the purpose of considering additional unexpected losses mentioned in Article
181(1)(h) of the CRR institutions may need to increase the LGD in-default over the downturn level.
Finally, as for the LGD for non-defaulted exposures, also the LGD in-default should include
appropriate MoC. In this context, irrespective of which of the approaches is used for LGD in-default
estimation, institutions should always be able to document:
•
the breakdown of the LGD in-default into its components: the ELBE and the add-on; and
•
the breakdown of the add-on into its components: the downturn adjustment, MoC, and,
where relevant, any component covering for additional unexpected losses during the
recovery period.
As the relation between ELBE and LGD in-default is crucial for the adequate determination of risk
weights, potential overrides have to be considered and applied consistently as well. It is therefore
specified in the draft GL that to the extent that the reason for overriding the ELBE also applies to the
LGD in-default an override of the LGD in-default should be triggered.
20
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Chapter 8: Application of risk parameters
Conservatism in the application of risk parameters
While the margin of conservatism described in Chapter 4 addresses any weaknesses in data or
methods in the process of model development and risk quantification, additional conservatism
referred to in Chapter 8 is meant to address any weaknesses in the implementation of the models
and application of the risk estimates to the currently existing exposures. These weaknesses may
include in particular missing or outdated information necessary for the rating assignment in
accordance with the model. Institutions should be able to detect and monitor these situations in
order to make sure that the risk is reflected correctly, including additional conservatism where
necessary. Some examples of the triggers that should lead to increased conservatism in the
application of risk parameters are included in Annex II.
Human judgement in the application of risk parameters
The proposed rules for the use of human judgement in the application of risk parameters are based
on Article 172(3) of the CRR which allows overriding both inputs and outputs of the assignment
process, both in the case of PD and LGD estimates, including ELBE and LGD in-default. In any case,
where institutions want to apply the overrides this should be based on an appropriate internal
framework to make sure that the weaknesses are identified consistently, that the overrides are
applied within certain limits and that they are appropriately justified, approved and monitored. As
the large number of overrides may indicate certain weaknesses of the model, institutions should
analyse these situations carefully, taking into account the reasons for overrides. Where necessary,
such analysis should result in the improvement of the model, for instance by including additional risk
drivers, increasing granularity of categorisation or changing the weights of risk drivers, or in the
improvement of data collection or data quality management processes.
Chapter 9: Re-development, re-estimation and re-calibration of
internal models
This Chapter is providing additional guidance for institutions to formulate triggers for redevelopment, re-estimation and re-calibration of risk parameters. In order to ensure that the
deterioration of the model performance is detected and addressed in a timely manner the draft GL
clarify what institutions should consider in their internal frameworks for annual reviews and what
should be the minimum scope of analysis that institutions conduct during this annual review.
Moreover, institutions are requested to define a cycle for a fundamental review of models depending
on the materiality of the models considered.
These reviews should contain an update of the development data set and re-estimation of model
components. An example framework specifying potential triggers for redevelopment and reestimation including specification of follow-up actions can be found in Annex IV.
21
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Chapter 10: Calculation of IRB shortfall or excess
Article 159 of the CRR requires institutions to calculate the difference between from one side credit
risk adjustments, additional value adjustments and other own funds reductions and from the other
side expected loss amounts for the purpose of own funds recognition. The GL refer to IRB excess
where this calculation result in a positive amount, i.e. where provisions are in excess over expected
loss, and to IRB shortfall where it results in a negative amount, i.e. there is a shortfall of provisions
given the expected loss. It has been clarified in Article 73(h) of the RTS on assessment methodology
that this difference should be calculated at an aggregate level separately for the portfolio of
defaulted exposures and the portfolio of exposures that are not in default. This separation is
necessary in order to ensure that the IRB excess resulting from the calculation performed for the
defaulted portfolio are not used to offset IRB shortfall resulting from the calculation performed for
the portfolio of exposures that are not in default as prescribed in Article 159 of the CRR. However,
the IRB excess from the overall non-defaulted portfolio may be used to cover any IRB shortfall from
the overall defaulted portfolio. It is furthermore clarified that if the calculation required by Article
159 of the CRR results in an IRB excess for both the defaulted and non-defaulted portfolio, the limit
for adding the overall IRB excess to TIER 2 capital set out in Article 62(d) of the CRR, i.e. up to 0.06%
of risk weighted exposure amounts, should be applied to the sum of the two IRB excess.
22
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Annex I – Example of list of triggers for appropriate adjustment and
additional MoC
Triggers for appropriate adjustment
and additional MoC
Potentially effected parameters
Category
Missing default trigger in historical
observations
DR, CureRate, RecoveryRate (EAD),…
A
Estimated date of default (late
detection default)
DR, CureRate, RecoveryRate (EAD),…
A
Length of historical default
observations (adjustments and MoC
according to Chapter 5.4.)
DR, CureRate, RecoveryRate (EAD),…
A
Changes in underwriting standards
(Diminished representativeness of
historical data to current exposure)
DR, CureRate, RecoveryRate (EAD),…
B
Changes in relevant processes
DR, CureRate, RecoveryRate (EAD), average
workout time, average internal and external
intensive care and workout costs,…
B
(Diminished representativeness of
historical data to current exposure)
Changes in legal environment
All
B
Missing collateral flags in historical
cash flows
RecoveryRate (EAD),…
B
Missing data in reference data set for
model development (risk drivers and
rating criteria)
DR per grade or pool, realized LGD per facility
grade
B
(Diminished representativeness of
historical data to current exposure)
Annex II – Example of list of triggers for conservatism in the
application of models
Triggers for appropriate adjustment and additional MoC
Potentially effected parameter
Missing data in current portfolio (application)
Single PD-Estimations, RWA
Missing update of financial statement
Single PD-Estimations, RWA
Missing re-rating in current portfolio (application)
Single PD-Estimations, RWA
Missing ratings:
Single PD-Estimations, RWA
Exposure wrongly without rating but in scope of a model
23
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Annex III: List of economic indicators to be taken into account for
determining the historical observation period for PD estimates for
particular exposure classes
Supervisory (sub-) exposure classes
Corporate
Retail
Economic indicators to be taken into account
in order to determine the historical
observation period for PD estimation
SMEs
GDP growth, unemployment rate, interest
rates, segmentation by industry sector,
default rate, credit losses
Other corporate
GDP growth, unemployment rate, interest
rates, segmentation by industry sector,
default rate, credit losses
SMEs
GDP growth, unemployment rate, industry
index, interest rates, branch, default rate,
credit losses
residential mortgages
House prices, GDP growth, unemployment
rate, interest rates, tax benefits, region,
default rate, credit losses
QRRE
GDP growth, unemployment rate, interest
rates, default rate, credit losses
other retail exposures
GDP growth, unemployment rate, consumer
price index, interest rates, default rate, credit
loss
Annex IV – Example of a list of triggers for re-development and reestimation
Area
Indicator
Trigger
Parameter(s) Action
The number of pools or
grades does not allow for
a meaningful
differentiation of risk
[Specification of
“meaningful”]
PD, LGD
Predictive Power
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The boundaries of pools
or grades do not allow
for a meaningful
differentiation of risk
[Specification of
“meaningful”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The pools or grades are
not composed of
[Specification of “not
PD where default
identification is on
[Short description of
action(s) to be taken with
24
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
homogenous groups of
obligors
homogenous”]
obligor level
reference to more detailed
documentation]
The pools or grades are
not composed of
homogenous groups of
facilities
[Specification of “not
homogenous”]
PD where default
identification is on
facility level and
LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The inclusion of current
data more recent than the
last calibration would
lead to materially
different model
outcomes
[Specification of
“materially”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The realised default rates
(either overall or per
rating grade) materially
differ from the expected
ones
[Specification of
“materially”]
PD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The realized LGDs
(either overall or in a risk
bucket) materially differ
from the long run
average expected one
[Specification of
“materially”]
LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The realized LGDs
(either overall or in a risk
bucket) materially differ
from the expected ones
under downturn
conditions
[Specification of
“materially”]
LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
A more recent downturn
period has been
identified
Representativeness &
Comparability
An undue concentration
has been identified in a
grade or pool
[Specification of
“undue concentration”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
[Other indicator(s)
identified]
[Specification(s) of
trigger(s) for other
indicator(s)]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The overall exposure
rated by the model has
significantly increased
compared to the
sample(s) used for
development, estimation
or calibration
[Specification of
“significantly
increased”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The size of the individual
facilities (ticket sizes)
has changed significantly
compared to the
sample(s) used for
development, estimation
[Specification of
“significantly
changed”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
25
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
or calibration
The maturity of the
individual facilities has
changed significantly
compared to the
sample(s) used for
development, estimation
or calibration
[Specification of
“significantly
changed”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The number and/or
exposure weighted
average PD of the
exposures has changed
significantly over time
[Specification of
“significantly
changed”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The shape of the
distributions of variables
relevant to the model
(design) has changed
compared to the
sample(s) used for
development, estimation
or calibration
[Specification of
“changed”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
A new type of
transaction, facility or
obligor has been
introduced in the scope
of the model without
requiring the competent
authorities’ approval
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
A new type of collateral
has been introduced in
the scope of the model
without requiring the
competent authorities’
approval
LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
A change in default
definition has taken
place since the time from
which the development
sample stems from,
which affects the
obligors or facilities in
the scope of the model
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The target
portfolio/lending
standards have changed
for the application
portfolio of the model
compared to the last
initial validation
[Specification of
“changed”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The work-out/recovery
process of defaulted
exposures has changed
materially compared to
the initial validation
[Specification of
“changed materially”]
LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
External market and
economic conditions or
[Specification of
PD, LGD
[Short description of
action(s) to be taken with
26
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Data Quality
other relevant
characteristics
surrounding the model
development have
changed materially
compared to the time
from which the
development sample
stems from.
“changed materially”]
[Other indicator(s)
identified]
[Specification(s) of
trigger(s) for other
indicator(s)]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Quality of data inputs is
unsatisfactory
[Specification of
“unsatisfactory”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Availability of data
inputs is unsatisfactory
[Specification of
“unsatisfactory”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Data available to the
institution is not entered
in the needed systems
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Data is no longer
available to the
institution
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Data is no longer
available in general
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Not all defaults are
properly recognized
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Not all defaults
recognised are
completely documented
and registered in all
appropriate and intended
IT systems
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Outdated input data is
used for the rating of
obligors or facilities
The number of technical
defaults is unsatisfactory
high
[Specification of
“outdated”]
[Specification of
“unsatisfactory”]
reference to more detailed
documentation]
27
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Discriminatory Power
Stability
[Other indicator(s)
identified]
[Specification(s) of
trigger(s) for other
indicator(s)]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The DP of a single risk
factor has either fallen
materially below the one
in the initial validation,
the last review or below
a fixed threshold
[Specification of
“fallen materially”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The DP of a scorecard
has either fallen
materially below the one
in the initial validation,
the last review or below
a fixed threshold
[Specification of
“fallen materially”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The DP of the overall
model has either fallen
materially below the one
in the initial validation,
the last review or below
a fixed threshold
[Specification of
“fallen materially”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
For application models,
the DP over the tenor of
the loan has either fallen
materially below the one
in the initial validation,
the last review or below
a fixed threshold
[Specification of
“fallen materially”]
PD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
[Other indicator(s)
identified]
[Specification(s) of
trigger(s) for other
indicator(s)]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The mean, standard
deviation, distribution or
extreme values of the
risk factors or other
relevant input parameters
of the model have
changed materially over
time
[Specification of
“changed materially”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The distribution of
reasons for obligor
defaults has changed
materially over time
[Specification of
“changed materially”]
PD where default
identification is on
obligor level
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The distribution of
reasons for facility
defaults has changed
materially over time
[Specification of
“changed materially”]
PD where default
identification is on
facility level and
LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The structure or number
of migrations between
grades or pools has
changed materially over
time
[Specification of
“changed materially”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
28
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Margin of Conservatism
Model Design
The PIT and TTC
characteristics of the
assignment to rating
grades or pools vary
significantly over time
[Specification of “vary
materially”]
PD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
[Other indicator(s)
identified]
[Specification(s) of
trigger(s) for other
indicator(s)]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
A model issue identified
which requires an MoC
does not have one
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Part of the MoC is in
place for a model issue
deemed to be rectified by
the body naming the
original issue
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
All or part of the MoC
calculation has not been
updated
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The overall MoC is
materially higher than
the one at the time of
model approval by the
competent authority or
than a fixed threshold
[Specification of
“materially higher”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
[Other indicator(s)
identified]
[Specification(s) of
trigger(s) for other
indicator(s)]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The model registry was
not complete or lacking
in quality
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Deficiencies in model
documentation have been
identified
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
A situation identified
during model
development under
which the model was
considered to perform
below expectations or
become inadequate has
or is expected to occur
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Materially better
modelling methods are
used in recent
developments
[Specification of
“materially better”]
29
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Use Test
[Other indicator(s)
identified]
[Specification(s) of
trigger(s) for other
indicator(s)]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The number of overrides
is too high or there is a
significant increase since
the last review
[Specification of “too
high” and “significant
increase”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Overrides negatively
affect the discriminatory
power of the model
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The rating outputs are
not used in the
institution’s internal
processes
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The re-ratings of
obligors or facilities are
not performed in a timely
manner
[Specification of “in a
timely manner”]
There are obligors or
facilities that are not
rated in the application
scope of the model
Benchmarking
IT Implementation
[Other indicator(s)
identified]
[Specification(s) of
trigger(s) for other
indicator(s)]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
There is a material
difference between the
external benchmarks and
the internal model
outputs
[Specification of
“material difference”]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
[Other indicator(s)
identified]
[Specification(s) of
trigger(s) for other
indicator(s)]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The IT implementation
of the model is not in
line with the
documentation
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
The data sources and
variables/risk factors
used for the development
are not properly
documented with regard
to their IT properties
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
Differences between the
data used by the model
development team and
the review responsible
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
30
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
unit have been identified
[Other indicator(s)
identified]
Compliance with
Regulatory Requirements
documentation]
[Specification(s) of
trigger(s) for other
indicator(s)]
The model is not
compliant with a new or
changed regulatory
requirement
[Other indicator(s)
identified]
[Specification(s) of
trigger(s) for other
indicator(s)]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
PD, LGD
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
[relevant
parameter(s)]
[Short description of
action(s) to be taken with
reference to more detailed
documentation]
31
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
3. Draft Guidelines
Draft Guidelines
on PD estimation, LGD estimation and
the treatment of defaulted exposures
32
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
1. Compliance and reporting
obligations
Status of these guidelines
1. This document contains guidelines issued pursuant to Article 16 of Regulation (EU) No
1093/2010 5. In accordance with Article 16(3) of Regulation (EU) No 1093/2010, competent
authorities and financial institutions must make every effort to comply with the guidelines.
2. Guidelines set the EBA view of appropriate supervisory practices within the European System
of Financial Supervision or of how Union law should be applied in a particular area.
Competent authorities as defined in Article 4(2) of Regulation (EU) No 1093/2010 to whom
guidelines apply should comply by incorporating them into their practices as appropriate (e.g.
by amending their legal framework or their supervisory processes), including where guidelines
are directed primarily at institutions.
Reporting requirements
3. According to Article 16(3) of Regulation (EU) No 1093/2010, competent authorities must
notify the EBA as to whether they comply or intend to comply with these guidelines, or
otherwise with reasons for non-compliance, by ([dd.mm.yyyy]). In the absence of any
notification by this deadline, competent authorities will be considered by the EBA to be noncompliant. Notifications should be sent by submitting the form available on the EBA website
to [email protected] with the reference ‘EBA/GL/201x/xx’. Notifications should be
submitted by persons with appropriate authority to report compliance on behalf of their
competent authorities. Any change in the status of compliance must also be reported to EBA.
4. Notifications will be published on the EBA website, in line with Article 16(3).
5
Regulation (EU) No 1093/2010 of the European Parliament and of the Council of 24 November 2010 establishing a
European Supervisory Authority (European Banking Authority), amending Decision No 716/2009/EC and repealing
Commission Decision 2009/78/EC, (OJ L 331, 15.12.2010, p.12).
33
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
2. Subject matter, scope and definitions
2.1 Subject matter
5. These guidelines specify the requirements for the estimation of Probability of Default (PD)
and Loss Given Default (LGD), including LGD for defaulted exposures (LGD in-default) and Best
Estimate of Expected Loss (ELBE) in accordance with Part Three, Title II, Chapter 3, Section 6 of
Regulation (EU) No 575/2013 as well as the application of Article 159 of that Regulation.
2.2 Scope of application
6. These guidelines apply in relation to the IRB Approach in accordance with Part Three, Title II,
Chapter 3 of Regulation (EU) No 575/2013. Where, for exposures other than retail, an
institution received permission to use the IRB Approach but has not received permission to
use own estimates of LGD in accordance with Article 143 and 151(8) to (9) of that Regulation,
the relevant parts of these guidelines apply, i.e. all parts excluding Chapters 6 and 7.
2.3 Addressees
7. These guidelines are addressed to competent authorities as defined in point (i) of Article 4(2)
of Regulation (EU) No 1093/2010 and to financial institutions as defined in Article 4(1) of
Regulation No 1093/2010.
2.4 Definitions
8. Unless otherwise specified, terms used and defined in Regulation (EU) No 575/2013 and
Directive (EU) 36/2013 have the same meaning in these guidelines. In addition, for the
purposes of these guidelines, the following definitions apply:
Risk parameters
One or all of the following: PD, LGD, ELBE and LGD in-default
PD model
All data and methods used as part of a rating system within the
meaning of Article 142(1) point (1) of Regulation (EU) No
575/2013, that relate to the differentiation and quantification
of own estimates of PD and which are used to assess the
default risk for each obligor covered by that model. A PD model
can contain several different methods for ranking the obligors
as well as different calibration segments.
Ranking method of a PD
model
The method used to rank the obligors with respect to the risk
of a default.
34
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Scoring method of a PD
model
The ranking method which assigns ordinal values (“scores”) to
rank obligors.
Calibration
For the purpose of quantifying the PD – the process of
developing a function (“Calibration Function”) that assigns PDs
to rating grades or pools in a manner that ensures that these
PD estimates correspond to the long run average default rates.
Calibration Segment
A subset of the range of application of the PD model which is
uniquely identified via the subset of obligors that are treated
with the same methods for the purpose of risk differentiation
and that are jointly calibrated.
LGD model
All data and methods used as part of a rating system within the
meaning of Article 142(1) point (1) of Regulation (EU) No
575/2013, that relate to the differentiation and quantification
of own estimates of LGD, LGD in-default and ELBE and which are
used to assess the level of loss in the case of default for each
facility covered by that model. An LGD model can contain
several different methods, especially with respect to different
types of collateral, which are combined to arrive at a LGD or
LGD in-default and ELBE for a given facility.
ELBE
Expected loss best estimate for defaulted exposures as referred
to in Article 153(1)(ii), Article 154(1)(i), Article 158(5) and
Article 181 (1)(h) of Regulation (EU) No 575/2013.
LGD in-default
Loss given default for defaulted exposures as referred to in
Article 153(1)(ii), Article 154(1)(i), Article 158(5) and Article 181
(1)(h) of Regulation (EU) No 575/2013.
IRB Shortfall
The difference, if negative, between, on the one hand, general
and specific credit risk adjustments, additional value
adjustments and other own funds reductions relating to these
exposures and, on the other hand, expected loss amount in
accordance with Article 159 of Regulation (EU) No 575/2013.
IRB Excess
The difference, if positive, between on the one hand, general
and specific credit risk adjustments, additional value
adjustments and other own funds reductions relating to these
exposures and, on the other hand, expected loss amount in
accordance with Article 159 of Regulation (EU) No 575/2013.
35
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
3. Implementation
3.1 Date of application
9. These guidelines apply from 1 January 2021. Institutions should incorporate the requirements
of these guidelines in their rating systems by that time, but competent authorities may
accelerate the timeline of this transition at their discretion.
3.2 First and ongoing application of the Guidelines
10. In order to apply these guidelines for the first time, and subsequently on a continuous basis,
institutions should assess and accordingly adjust, where necessary, their rating systems so
that the estimates of risk parameters reflect the requirements specified in these guidelines as
well as those specified in Commission Delegated Regulation xxx/xxxx [RTS on IRB assessment
methodology].
11. The changes referred to in paragraph 10, which are applied to the rating systems as a result of
the application of these guidelines, are required to be verified by the internal validation
function and classified according to Commission Delegated Regulation (EU) No 529/2014 6,
and, depending on this classification, they are required to be notified or approved by the
relevant competent authority.
12. Institutions which need to obtain prior permission from competent authorities in accordance
with Article 143 of Regulation (EU) No 575/2013 and Regulation (EU) No 529/2014 in order to
incorporate these guidelines for the first time by the deadline referred to in paragraph 9,
should agree with their competent authorities the final deadline for submitting the
application for the approval of changes in the rating systems.
3.3 Repeal
13. Sections [XXX] of the CEBS Guidelines on the implementation, validation and assessment of
Advanced Measurement (AMA) and Internal Ratings Based (IRB) Approaches (GL10) published
on 4 April 2006 are repealed with effect from 1 January 2021.
6
OJ L 148, 20.5.2014, p. 36.
36
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
4. General estimation requirements
4.1 Segmentation principles
14. A rating system as referred to in point (1) of Article 142 should cover exposures where the
obligors or facilities show common characteristics of credit-worthiness and fundamentally
comparable availability of credit related information.
15. Exposures covered by the same rating system should be treated similarly by the institution in
terms of risk management, decision making and credit approval process and should be
assigned to a common obligor rating scale in accordance with Article 170(1) point (b) of
Regulation (EU) No 575/2013 where applicable.
16. For the purpose of estimation of different risk parameters within a rating system, institutions
should ensure consistency with respect to the applicable definition of default and of the
default observations considered.
4.2 Data requirements
17. In order for the data used as inputs into the model to be accurate as required by Article
174(b) of Regulation (EU) No 575/2013 in the course of assigning exposures to obligors or
facility grades or pools, data should be sufficiently precise to avoid material distortion of the
outcome.
18. In order for the data used as inputs into the model to be complete as required by Article
174(b) of Regulation (EU) No 575/2013, these data should provide comprehensive
information for the institution including data for all relevant business lines and all relevant
variables, and institutions should attempt to minimise the occurrence of missing data.
19. In order for the data used as inputs into the model to be appropriate, as required by Article
174(b) of Regulation (EU) No 575/2013, data should not contain biases which make them
unfit-for-purpose.
20. For the purpose of Article 76 of Commission Delegated Regulation xxx/xxxx [RTS on IRB
assessment methodology] institutions should specify internal policies, standards and
procedures for data collection, storage, migration, actualisation and use, with such
characteristics so as to ensure regular updating and correcting of the data where necessary.
21. The process for vetting data which includes an assessment of the accuracy, completeness and
appropriateness of the data, as required by Article 40 of Commission Delegated Regulation
xxx/xxxx [RTS on IRB assessment methodology] should include in particular all of the
following:
37
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(a) the assessment of reliability and quality of the internal and external data sources and the
range of data obtained from those sources, as well as the time period the sources cover;
(b) the data merging, where the model is fed with data from multiple data sources;
(c) the rationale and scale of data exclusions broken down by reason for exclusion, using
statistics of the share of total data covered by each exclusion, where certain data were
excluded from the model development sample;
(d) the procedures for dealing with erroneous and missing data and treatment of outliers and
categorical data, and the procedures for ensuring that, where there has been a change in
the type of categorization, this did not lead to decreased data quality or structural breaks
in the data;
(e) the data transformation, including the standardization and other functional
transformations and the procedures for ensuring the appropriateness of those
transformations in terms of the risk of model overfitting.
4.3 Human judgement in model development
22. In order for institutions to complement their statistical models with human judgement, as
referred to in Articles 172(3), 174(b), 174(e), 175(4), 179(1)(a), 180(1)(d) of Regulation (EU)
No 575/2013, they should meet all of the following conditions:
(a) they should in particular assess the modelling assumptions and whether the selected risk
drivers contribute to the risk assessment in line with their economic meaning;
(b) they should ensure that any form of human judgement is properly justified and should
analyse the impact of the human judgement on the performance of the model.;
(c) they should document the application of human judgement in the model , include at least
the criteria for the assessment, rationale, assumptions, experts involved and description
of the process.
4.4 Margin of conservatism (‘MoC’)
23. In relation to the requirement that institutions should add a margin of conservatism (‘MoC’)
that is related to the expected range of estimation errors as required by Article 179(1) point
(f) of Regulation (EU) No 575/2013 and Article 180(1) point (e), institutions should implement
a framework, that consists of the phases specified in sections 4.4.1 to 4.4.4.
4.4.1
Identification of deficiencies
24. Institutions should have a robust process for identifying all deficiencies, including data errors
and any uncertainties that lead to estimation errors, and for classifying them in the following
categories:
38
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(a) Category A: Expected range of estimation errors due to data deficiencies;
(b) Category B: Expected range of estimation errors due to diminished representativeness of
historical observations;
(c) Category C: General estimation errors including errors stemming from methodological
deficiencies;
(d) Category D: Other uncertainties.
25. For the purposes of applying the MoC during the phases of model development, estimation
and calibration institutions should consider:
(a) for the errors classified under Category A as referred to in paragraph 24 at least the
following triggers:
(i) missing or materially changed default triggers in historical observations;
(ii) missing estimated date of default, leading to late default detection;
(iii) missing or outdated rating information used for the purpose of calculation of default
rate per grade or pool;
(iv) missing or inaccurate information on the source of cash flows;
(v) missing, inaccurate or outdated data on risk drivers and rating criteria;
(vi) missing or inaccurate data for the calculation of economic loss;
(b) for the errors classified under Category B as referred to in paragraph 24, at least the
following triggers:
(i)
diminished representativeness of the historical observations due to the changes in
the definition of default
(ii) diminished representativeness of the historical observations due to the use of
external data
(iii) diminished representativeness of the historical observations due to changed
underwriting standards or recovery policies;
(iv) diminished representativeness of the historical observations to the current portfolio
in terms of the distribution of risk drivers;
(c) for the errors classified under Category C as referred to in paragraph 24, methodological
errors not yet rectified, including:
(i) the rank order estimation error ;
39
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(ii) estimation error in the calibration;
(d) for the errors classified under Category D as referred to in paragraph 24, at least the
following triggers:
(i) changes in the legal environment not covered by the errors included under Category
B referred to in paragraph 24;
(ii) changes in the relevant processes not covered by the errors included under Category
B referred to in paragraph 24;
(iii) estimation error in the long-run averages due to necessary adjustments to comply
with Article 179(1)(d), Article 49(3) to (5) and Article 53 of Commission Delegated
Regulation xxx/xxxx [RTS on IRB assessment methodology];
4.4.2
Quantification of estimation errors
26. In order to overcome estimation errors in PD and LGD estimates stemming from the
categories of deficiencies A, B or D, institutions should apply adequate methodologies for
correcting the identified errors (‘appropriate adjustment’). Institutions should ensure that
the appropriate adjustment results in a more accurate estimate of the risk parameter, where
this adjustment can have both positive and negative effect on the risk parameter.
27. Where such appropriate adjustments are used institutions should apply a MoC to account for
the additional estimation error associated with these adjustments. The MoC related to the
economic adjustment should be proportionate to the impact of the adjustment on the risk
parameter.
28. Institutions should also apply a MoC to address any errors that have not been corrected via
appropriate adjustment and any identified uncertainties. Institutions should ensure that the
impact of the MoC does not ever result in lowering PDs or LGDs.
29. Institutions should assess the MoC at the level it is identified but they should reflect and
report it with respect to the final risk parameter estimate used for own funds requirements.
30. Any occurrence of any of the triggers referred to in paragraph 25 should result in the
application of a MoC. Where more than one trigger occurs, a higher aggregate MoC should be
applied. The MoC related to each trigger should be proportionate to the estimation error in
the estimated parameter that results from the identified deficiency. Institutions should
quantify the estimation error that results from the identified deficiency in order to justify the
level of MoC. Institutions should quantify the appropriate adjustment and MoC as defined in
paragraphs 26 to 29 at least for every calibration segment.
31. Institutions should provide for a customable IT implementation solution, which ensures that
MoC can be implemented in a timely manner.
40
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
32. Institutions should consider the overall impact of the identified deficiencies and the resulting
MoC on the soundness of the model and ensure that capital requirements are not distorted
due to the necessity for excessive adjustments.
4.4.3
Monitoring
33. Institutions should regularly monitor the levels of the appropriate adjustments and MoC. The
adoption of a MoC by institutions should not replace the need to address the causes of errors
or uncertainties and to correct the models to ensure their full compliance with the
requirements of the Regulation (EU) No 575/2013. Following its assessment, institutions
should develop a plan to rectify the data and methodological deficiencies and reduce the
estimation errors within a reasonable timeline, taking into consideration the materiality of
the estimation error and the materiality of the rating system.
34. When reviewing the levels of MoC institutions should ensure all of the following:
(a) that the MoC stemming from Category A, B and D as referred to in paragraph 24 is
reduced over time;
(b) that the MoC stemming from Categories C as referred to in paragraph 24 is eliminated
after the error is rectified in all parts of the rating system that were affected.
4.4.4
Documentation
35. For each rating system, the MoC applied should be documented in the relevant model
documentation and methodology manuals. The documentation should at least contain:
(a) a complete list of all potential and identified deficiencies and the potentially affected
model components or risk parameters,
(b) a description of the methods used to apply appropriate adjustments to rectify the data
and methodological errors, where relevant;
(c) a description of the methods of addressing the deficiencies, including errors and
uncertainties, via the application of an MoC;
(d) the category under which these errors and uncertainties are classified, as referred to in
paragraph 24.
Explanatory Box for consultation purposes:
The proposed requirements are based on the assumption that the identified deficiencies in data
or methods should be addressed via appropriate adjustment (where applicable) and margin of
conservatism:
41
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
-
Appropriate adjustment consists in rectifying the identified errors, for instance missing data
points are filled in with the most probable information or the inaccuracies in data are
corrected. The objective of the appropriate adjustment is to achieve the possibly most
accurate estimates. The impact of the appropriate adjustment may be either positive or
negative depending on the character of the specified deficiency.
-
Margin of conservatism aims at addressing all errors that cannot be rectified through
appropriate adjustment and any other uncertainties related to the estimation of risk
parameters. In addition to that, additional MoC has to address the additional uncertainty
related to the application of appropriate adjustment, as the adjustment is only an
approximation of the actual events or characteristics. MoC cannot lead to decrease in the
estimates
It is required by Article 179(1)(f) of Regulation (EU) 575/2013 that margin of conservatism, related
to the expected range of estimation errors, shall be added to the estimates of risk parameters. It
is therefore clarified in the draft Guidelines that institutions should be able to calculate and report
the exact impact of the MoC at the level of risk parameters even if it is identified at earlier stages
of the estimation.
The Guidelines do not prescribe any specific method for the quantification of MoC as the
appropriate approach will depend on the character of the deficiency and the available data.
However, institutions should keep in mind that model aspects that appear conservative in one
model may not be truly conservative compared with alternative methods. For example, simply
picking an extreme point on a given modelled distribution may not be conservative if the
distribution was misestimated or misspecified in the first place. Furthermore, initially
conservative assumptions may not remain conservative over time. Therefore, it is expected that
the methods applied to derive the MoC will be regularly revised in order to ensure that the effect
on the risk parameters is adequate and proportionate to the estimation error related to the
identified deficiencies.
Question 4.1: Do you agree with the proposed requirement with regard to the application of
appropriate adjustments and margin of conservatism? Do you have any operational concern
with respect to the proposed categorization?
42
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
5. PD estimation
5.1 General requirements specific to PD estimation
36. In order for the input variables into a PD-Model to form a reasonable and effective basis for
the resulting predictions, in accordance with Article 174(a) of Regulation (EU) No 575/2013,
the PD-Model should cover exposures where the obligors show common drivers of risk and
fundamentally comparable availability of credit-related information.
37. Exposures covered by one PD Model should be managed homogeneously by the institution in
terms of risk management, decision making and credit approval process.
38. For the purpose of assigning obligors to an obligor grade as part of the credit approval process
in accordance with Article 172(1)(a) of Regulation (EU) 575/2013 as well as for the purpose of
annual review of those assignments, in accordance with Article 173(1)(b) and (2) of that
Regulation, institutions should ensure that each and every natural or legal person that
represents an IRB exposure is rated by the institution with the model approved to be used on
and fitting to the single original obligor, including where there is unfunded credit protection
as referred to in Article 161(3) of that Regulation.
39. In the application of the rating model for PD, institutions should ensure that where an
institution receives new information with respect to a relevant risk driver or rating criterion,
this information is taken into account in the rating calculation in a timely manner. In
particular,
(a) if this information has to be updated in the relevant IT systems, a resulting review of the
rating assignment should not be made later than 3 months after the information becomes
available.
(b) in case of retail exposures where the new information is available in other IT-Systems,
these should be taken into account in the next rating review.
(c) in case of a default of an obligor the PD of the obligor should be set to 1 in a timely
manner in all relevant IT-Systems;
5.2 Data requirements specific to PD estimation
5.2.1
Data requirements for default rate calculation
40. For the purpose of calculating the default rate defined in Article 4(78) of Regulation (EU) No
575/2013, institutions should ensure the completeness of the quantitative and qualitative
data and other information in relation to the denominator and numerator as outlined in
paragraph 48 and 49 used for the calculation of the observed average default rate, and more
43
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
in particular that at least the following data is properly stored and available for the relevant
observation period referred to in paragraphs 59 to 61:
(a) the criteria for identifying the relevant type of exposures covered by the PD model under
consideration;
(b) the risk drivers used for risk differentiation according to the rating method considered for
the purpose of calculating the long-run average default rate per grade or pool. Where a
newly relevant risk driver has been included into the model for which no historical data is
available, an appropriate adjustment and MoC should be applied. As outlined in
paragraph 34(a) institutions should make efforts to minimise, over time, the share of
ratings where a considered risk driver is missing;
(c) all identification numbers (‘IDs’) relevant for default rate calculation, in particular where
the ID has changed due to a restructuring.
41. Exclusion of observations from the default rate calculation should be done exclusively in the
following two situations:
(a) obligors wrongly included in the data set of defaults and which did not default with the
meaning of default as specified in the Guidelines on the definition of default of an obligor
as referred to in Article 178 (7) of Regulation (EU) No 575/2013 should not be included in
the numerator;
(b) obligors wrongly assigned to the considered rating model , because they are not falling in
the range of application of that rating model, should be excluded from the numerator and
the denominator.
42. Institutions should document all data cleansing in accordance with Article 32 (3) (b) of
Commission Delegated Regulation xxx/xxxx [RTS on IRB assessment methodology], with
respect to the default rate calculation and in particular include:
(d) for non-retail rating models, a list of all observations that were excluded according to the
previous paragraph, with a case by case justification;
(e) for retail rating models, information on the reasons and quantity of exclusions of
observations made in accordance with the previous paragraph.
43. Institutions should ensure consistency between the data sets deriving from different data
sources used for the calculation of the observed average default rate, particularly with regard
to the default definition and the treatment of multiple defaults.
5.2.2
Data requirements for the reference data set for the purpose of model
development
44. Institutions should provide for sound processes and sophisticated methods so as to either
take into account or compensate with the addition of MoC, of all of the following when
constructing the reference data set for the purpose of model development:
44
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(a) unsatisfactory quality of data;
(b) lack of homogeneous pools of exposures;
(c) changes in business processes, the economic or legal environment;
(d) other factors relating to the quality of data that may affect the performance of the PDmodel.
45. With regard to the representativeness requirement of Article 174 point (c) of Regulation (EU)
No 575/2013 institutions should ensure all of the following:
(a) that representativeness applies to both statistical models and other mechanical methods
used to assign exposures to grades or pools, and to statistical default prediction models
generating default probability estimates for individual obligors or facilities;
(b) that the same approach with regard to representativeness in the sense of Article 174
point (c) of Regulation (EU) No 575/2013 applies regardless of whether the data is
internal, external, or consists of pooled data sets, or a combination of the above;
(c) that, for the purpose of analysing the representativeness of the sample with respect to
the current population covered by the considered model, all quantitative and qualitative
obligor and facility characteristics that could relate to default for PD estimation should be
anaysed, and more in particular all of the following:
(i) the comparability of the underwriting and recovery standards with the ones applied
at the time of the reference data set used for the modelling;
(ii) where there are more than one sets of the same obligor or facility characteristic, a
mapping from one set of characteristics to the other should be applied.
(iii) the distribution of the current population and the sample according to the key
characteristics and the level and range of these key characteristics. Material observed
differences in the key characteristics should be avoided, for example by using another
sample;
(d) that where applicable, statistical methodologies such as cluster analysis or related
techniques should be used to demonstrate representativeness;
(e) that the definition of default is consistent over time in the data used for the modelling,
and more in particular:
(i) that adjustments have been made to achieve consistency with the current default
definition where the default definition has been changed during the observation
period;
45
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(ii) that adequate measures have been adopted by the institution, where the model
covers exposures in several jurisdictions having or having had different default
definitions;
(iii) that the default definition used for the purposes of model development does not
have a negative impact on the structure and performance of the rating model, in
terms of risk differentiation and predictive power, where this definition is different
from the definition of default used by the institution in accordance with Article 178 of
Regulation (EU) 575/2013;
(f) that external data or data pooled across institutions for the institution’s exposures,
products and risk profile are relevant and adequate, where such data is used in the model
development.
46. Representativeness for the assignment of obligors or facilities does not require that the
proportion of defaulted and non-defaulted exposures in the development data set be equal
to the proportion of defaulted and non-defaulted exposures in the institution’s respective
portfolio.
47. It should be ensured that the reference data set contains the values of the risk drivers for
appropriate points in time. These points in time may vary between different risk drivers. In
the selection of appropriate points in time the institution should take into account the
dynamics of the risk drivers throughout the whole period in which an obligor was in the
portfolio and, in the case of a default, especially throughout the year prior default. For
example an institution may base its reference data set on monthly snapshots of the portfolio
considered over the whole historical observation period.
5.3 Observed default rates
5.3.1
Calculation of the one-year default rate
48. For the purpose of calculating the one-year default rate as referred to in Article 4 (1) point
(78) of Regulation (EU) No 575/2013, both of the following should apply:
(a) the denominator should consist of the number of non-defaulted obligors observed at the
beginning of the one-year observation period with any credit obligation. In this context a
credit obligation refers to any amount of principal, interest and fees as well as to any offbalance sheet items including guarantees.
(b) the numerator should include all obligors considered in the denominator with at least one
default event during the one-year observation period.
49. Where the one-year-default-rate is calculated by rating grade or pool the denominator should
refer to all obligors assigned to a rating grade or pool at the beginning of the observation
period, taking into account overrides, but excluding any substitution effects due to credit risk
mitigation, as well as any ex-post conservative adjustments.
46
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
50. Institutions should calculate the one-year default rate also for the subset of obligors that did
not have a rating at the start of the relevant observation period but were in the range of
application of the model under consideration, even if these obligors were assigned to a rating
grade or pool in a conservative manner for the purpose of calculation of capital requirements
(‘missing ratings’). Obligors whose ratings are based on missing or partly missing information
or where the rating is outdated but still deemed valid by the institution should not be
considered as missing ratings.
51. For the avoidance of doubt with regard to paragraphs 48 to 50 an obligor has to be included
into the denominator, and numerator as well, if relevant, also in case of a migration to a
different rating grade, pool or rating model, rating system or approach to calculation of
capital requirements within the observation period or where the corresponding credit
obligations were sold during the observation period. Institutions should analyse whether such
migrations bias the default rate and if so reflect this in an appropriate adjustment and
consider such bias in their determination of an appropriate margin of conservatism.
52. In cases where there is a significant proportion of customers carrying multiple facilities within
a considered Retail rating system and the institution identifies defaults at the level of an
individual credit facility institutions should ensure that the estimates are not biased due to
the multiple facilities.
53. In order to monitor the appropriateness of the PD estimates, institutions should calculate the
one-year default rates at least quarterly.
Question 5.1: Do you see any operational limitations with respect to the monitoring
requirement proposed in paragraph 53?
5.3.2
Calculation of the observed average default rate
54. The observed average of one-year default rates (‘observed average default rate’) should be
calculated per rating grade or pool and should additionally be calculated for the portfolio
covered with the according PD Model as well as for any relevant calibration segment.
55. Institutions should document the considerations for the chosen approach to calculating the
observed average default rate. This documentation should include :
(a) for institutions using overlapping one-year time windows, an analysis of a potentially
significant bias that occurs due to implicit down weighting of defaults that occurred in the
first and last time slice;
(b) for institutions using non-overlapping one-year time windows, an analysis of a potentially
significant bias due to seasonal effect related to the chosen calculation date.
56. For the purpose of choosing an appropriate approach for calculating the observed average
default rate institutions should analyse among others at least both of the following:
47
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(a) the share of short term and terminated contracts that cannot be observed during the
whole observation period, which might bias the default rate;
(b) the possible bias due to specific reporting dates chosen.
57. For the purpose of paragraphs 55 and 56, institutions should apply an economic adjustment
and an appropriate MoC, where applicable.
58. For the purpose of calculating the observed average default rate the defaults are not to be
weighted but each to be counted as 1.
Explanatory box for consultation purposes:
Observed average default rate:
1)
Non-overlapping windows
Let M be the set of relevant years for calculating the observed average default rate and let DR(i) denote the
one-year default rate in year i. The observed average default rate ODR when calculated via an arithmetic
average is given as
𝑂𝑂𝑂𝑂𝑂𝑂 =
1
𝑁𝑁
∑𝑁𝑁
𝑖𝑖=1 𝐷𝐷𝐷𝐷(𝑖𝑖), where 𝑖𝑖 = 1, … . , 𝑁𝑁 denote the elements of M, which are usually assumed to be
consecutive, non-overlapping years. Thus the default rates 𝐷𝐷𝐷𝐷(1), … , 𝐷𝐷𝐷𝐷 (𝑁𝑁) are calculated yearly with
respect to a certain reference date 𝑇𝑇𝑖𝑖 .
The observed average default rate which is calculated according to this method could be biased due to the
choice of fixed reference dates 𝑇𝑇𝑖𝑖 . This means in particular that the representativeness of these default rates
could be questioned since choosing certain dates could lower the observed average default rate und thus
lead to lower capital requirements.
2)
Overlapping windows
Assume that an institution calculates the one-year default rate with a monthly frequency. Now for
calculating the observed average default rate using overlapping windows the following applies:
𝑂𝑂𝑂𝑂𝑂𝑂 =
1
𝑗𝑗
∑𝑗𝑗𝑖𝑖=1 𝐷𝐷𝐷𝐷(𝑖𝑖), where 𝑖𝑖 = 1 … 𝑗𝑗 = 12𝑁𝑁 denote the 12𝑁𝑁month of N considered consecutive years.
The observed average default rate which is calculated according to this method could be biased due to the
implicit down weighting of the defaults in the initial and latest one-year time windows of the historical
observation period, i.e. defaults which occur in the middle of the observation period are counted 12 times
whereas defaults which occur in the first or last year are counted only 1-11 times.
Question 5.2: Do you agree with the proposed policy for calculating observed average default
rates? How do you treat short term contracts in this regard?
5.4 Long-run average default rate
59. For the purpose of determining the historical observation period referred to in Article
180(1)(h) and 180(2)(e) of Regulation (EU) No 575/2013, additional observations to the most
48
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
recent 5 years, at the time of model calibration, should be considered as relevant when these
observations are representative of the likely range of variability of default rates of that type of
exposures as referred to in Article 49(3) of Commission Delegated Regulation xxx/xxxx [RTS on
IRB assessment methodology].
60. When the historical observation period as referred to in paragraph 59 is representative of the
likely range of variability of default rates, then the long-run average default rate should be
computed as the observed average of the one-year default rates in that period.
61. For the purpose of assessing the representativeness of historical observation period as
referred to in paragraph 59 for the likely range of variability of one-year-default rates,
institutions should take into account all of the following:
(a) the variability of all observed one-year-default rates;
(b) the existence or lack of one-year default rates relating to downturn periods as reflected
by economic indicators that are relevant for the considered type of exposure within the
historical observation period;
(c) significant changes in the economic, legal or business environment within the historical
observation period.
62. In case the historical observation period is not representative of the likely range of variability
of one year default rates in order to comply with Article 49(4) of Commission Delegated
Regulation xxx/xxxx [RTS on IRB assessment methodology] the average of observed one year
default rates should be adjusted in order to estimate a long-run average default rate, in
particular where no downturn period is included in the historical observation period.
63. In case that the long-run average default rate does not equal the average of all observed one
year default rates, institutions should compare their adjusted long-run average default rates
to the maximum between:
(a) the observed average of the one-year default rates of the most recent 5 years and
(b) the observed average of all available one-year default rates
and where the adjusted long-run average default rate is lower than that maximum institutions
should justify the direction and magnitude of the adjustment, including the adequacy of the
considered margin of conservatism, where applicable.
Explanatory box for consultation purposes:
Relevant observation period:
The current approach requires that institutions analyse, for the purpose of determining the
relevant historical observation period, whether a downturn period is included in the available
observation period or not. For this purpose institutions should analyse the sensitivity of the level
49
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
and volatility of observed defaults within a rating system to economic indicators, considering
economy as a whole, as well as addressing more specific (e.g., industry) cycles. A non-exhaustive
list of economic indicators that should be taken into account for this purpose is attached in Annex
III of the background and rationale section.
Moreover, in case a downturn period is contained in the available observation period, institutions
may in order to estimate the long-run average default rate estimate an adjustment to the
observed average default rate if it does not reflect an appropriate mix of favorable and
unfavorable economic conditions, which would be the case for example if the available
observation period would cover mainly downturn or mainly good years. Institutions should in any
case adjust the observed average default rate if the default rates are sensitive to the economic
indicators and no downturn is contained in the available observation period.
Other adjustments may, however in specific cases, be necessary in order to comply with Article
179(1)(d) of Regulation (EU) No 575/2013. It should be noted that the benchmark also applies in
this cases. Including such specific cases into the GL has been discussed as well, in particular where
downward adjustments to the average of observed one-year default rates would be justified. In
this regard, sustained tightening of underwriting standards, changed relevant legislation, changed
business environment, mergers & acquisitions and changes of internal processes have been
mentioned.
Question 5.3: Are the requirements on determining the relevant historical observation periods
sufficiently clear? Which adjustments (downward or upward), and due to which reasons, are
currently applied to the average of observed default rates in order to estimate the long-run
average default rate? If possible, please order those adjustments by materiality in terms of
RWA.
5.5 PD estimation methodologies
5.5.1
Risk drivers and rating criteria
64. In the process of selecting risk drivers and rating criteria, institution should consider a broad
scope of information, including obligor characteristics, for example sector and geographic
location for corporates, financial statements as well as trend and behavioral information.
Relevant trend information could for example include growing or shrinking sales or profit
margin and relevant behavioural information could for example include the use of overdrafts
if applicable.
65. Institutions should ensure that for the purpose of selecting risk drivers and rating criteria the
relevant business experts are consulted with respect to the business rationale and risk
contribution of the considered risk drivers and rating criteria.
66. Institutions should ensure that the loss of information value over time for generally static
information, for instance the information on obligor characteristics at the time of the loan
application, is appropriately reflected. It has to be ensured that the model estimates the
50
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
proper level of risk with respect to all relevant, currently available and most up-to-date
information and that an appropriate MoC is applied where a higher degree of uncertainty is
probable due to the lack of up-to-date information. In particular the model or the assignment
process should provide adequate and conservative adjustment in both of the following
situations:
(a) in accordance with Article 24(1)(g) of Commission Delegated Regulation xxx/xxxx [RTS on
IRB assessment methodology], in case of financial statements older than 24 months
where information stemming from these financial statements are relevant risk drivers
(b) in case of credit bureau information older than 24 month, if still relevant at that point in
time, where credit bureau information is a relevant risk driver.
67. Institutions should ensure that the risk drivers and rating criteria are used consistently, in
particular with respect to the considered time horizon, in model development, model
calibration and model application.
5.5.2
Ratings in PD estimation
68. Institutions should have clear policies specifying the triggers resulting from the contractual
relation between a third counterparty (‘connected client’) and the considered obligor that
lead to each of the following outcomes:
(a) triggers resulting in the rating of that connected client being transferred to a considered
obligor due to CRM substitution (‘rating transfer’), according to Article 161(3) of
Regulation (EU) No 575/2013;
(b) triggers resulting in a rating of a connected client being taken into account either as
indication for an override of the individual PD estimate of the considered obligor;
(c) triggers resulting in a rating of a connected client serving as input to the PD model (‘a
support’).
69. In the course of establishing the policies referred to in the previous sub-paragraph,
institutions should take into account paragraphs 70 to 74.
70. In order for an internal or external rating of connected clients to be incorporated into a
statistical model, the rating should comply with all of the following:
(a) it should fulfil all the requirements for relevant risk drivers laid down in section 5.5.1;
(b) the weighting in the statistical model should be purely statistically based;
(c) institutions should ensure that other relevant obligor and transaction risk characteristics
are properly reflected in the model in accordance with Article 170(1) point (a) and Article
51
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
170(3) point (a) of Regulation (EU) No 575/2013 and that no material biases are
introduced by a high weighting of the internal or external rating information.
71. An internal IRB rating for a connected client may be incorporated in the non-statistical part of
the PD model or through the use of overrides, if not already incorporated in the statistical
part.
72. A rating transfer should not change the assignment of exposures to exposure classes, rating
systems or models, but should only affect the assignment to grades or pools. Rating transfers
should be set up in such a way that any changes to a rating of a connected obligor which is
material information on the obligor or exposure with regard to Article 173(1) point (b) of
Regulation (EU) No 575/2013 is reflected in all influenced ratings in a timely manner.
73. An institution’s policy should prevent inappropriate double counting of a contractual relation
to a connected client or group of connected clients.
74. The possible support of one obligor to another should be seen as diminishing the free
financial strength of the supporting obligor, including the strength to repay all obligations to
the institution in full without recourse, irrespective of the rating transfer method chosen. This
should be reflected in the rating of the supporting obligor.
5.5.3
Design of grades or pools
75. Depending on the methods and drivers used to assign exposures to risk grades or pools,
changes in the portfolio’s default rate caused by changes in economic conditions will be
reflected through a combination of:
(a) migrations across risk grades;
(b) changes in the yearly default rates of each grade.
76. Where the rating assignment process is highly sensitive to the economic conditions, grades
assignment will change significantly, while default rates of each grade will re-main relatively
stable. In contrast, when the assignment is less sensitive to the economic conditions, the
yearly default rates per grade component will capture the cyclicality of the global default rate.
77. Institutions should analyse the appropriateness of the philosophy underlying the grade or
pool assignment in terms of how institutions assign exposures, obligors or facilities to ‘risk
buckets’ according to appropriate risk drivers;
78. Institutions should decide the philosophy underlying the grade or pool assignment, and
specifically the risk drivers. However,
(a) the choice of rating philosophy should be applied consistently over time;
52
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(b) Institutions should assess whether the method used to quantify the risk parameter is
adequate for the philosophy underlying the grade or pool assignment and understand the
characteristics and dynamics, of the ratings and of the risk parameter estimates that
result from the method used.
(c) Institutions should assess the adequacy of the resulting characteristics and dynamics, of
the ratings and risk parameter estimates that result from the method used, with regard to
their different uses and should understand their impact on the dynamics and volatility of
capital requirements.
(d) The rating philosophy must also be taken into account for back testing purposes.
Sensitive philosophies tend to estimate PDs which are better predictors of each year’s DR.
On the other hand, more insensitive philosophies tend to estimate PDs which are closer
to the average PD across the different states of the economy, but that differ from
observed DRs in years where the state of the economy is above or below its average.
Deviations between observed default rates and the average will hence be more frequent
in rating system less sensitive to the cycle. On the contrary, migrations among grades will
be more frequent in rating system more sensitive to the cycle. These patterns have to be
taken into account when analysing back-testing results. They shall also be accounted for
in benchmarking analysis.
79. When an institution uses different rating systems characterised by different philosophies,
care should be taken in the use of information, either for rating assignments or estimates,
from another, internal or external, rating system that has a different rating philosophy. An
example is the use of rating information or default experience obtained from rating agencies.
When an institution uses different rating systems with different characteristics, for example
different philosophies, levels of objectivity, accuracy, stability, or conservatism, it should
ensure that they have an appropriate level of consistency and/or that the differences
between them are well understood. This understanding should at least enable the institution
to define an appropriate way to combine/aggregate the information produced by the
different rating systems when this is necessary. The assumptions and potential inaccuracies
arising from such a combination/aggregation should be fully understood.
Explanatory box for consultation purposes:
RWA variability stemming from different rating philosophies
Different rating philosophies is listed as a potential driver of unjustified RWA variability across
institutions in several analysis conducted e.g. by the IIF and EBA. It is, however due to a lack in
common terminology and measures to specify different rating methodologies up to now not
possible to prove or disprove that the differences in rating philosophy necessarily lead to RWA
variability.
The analysis of migration matrices is probably the most straightforward tool in order to observe
the dynamics of a rating system. The migration matrix of a model which is insensitive to economic
53
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
conditions would tend to be symmetrical independently of the state of the economy, while that
of a sensitive model will tend to be biased toward upgrades or downgrades depending on the
state of the economy.
The evolution of default rates at grade level and portfolio level can also be used in order to
analyse rating dynamics: sensitive ratings would show significantly less dynamics due to the
economic environment at grades’ DRs than would show at portfolio level, while insensitive ratings
would tend to show similar dynamics both at portfolio and grade level. Thus the qualitative survey
aims to gather information on this topic and asks a number of related questions.
Question 5.4: How do you take economic conditions into account in the design of your rating
systems, in particular in terms of:
a. definition of risk drivers,
b. definition of the number of grades
c. definition of the long-run average of default rates?
Question 5.5: Do you have processes in place to monitor the rating philosophy over time? If yes,
please describe them.
Question 5.6: Do you have different rating philosophy approaches to different types of
exposures? If yes, please describe them.
Question 5.7: Would you expect that benchmarks for number of pools and grades and
maximum PD levels (e.g. for exposures that are not sensitive to the economic cycle) could
reduce unjustified variability?
5.5.4
Calibration
80. Institutions should have sound and well-defined processes in place to ensure that accurate
and robust PD estimates are assigned to grades or pools of obligors or facilities. Institutions
should ensure a sound calibration by including the following in their calibration process:
(a) quantitative calibration tests by rating grade or pool;
(b) where applicable, quantitative calibration tests on calibration segment level;
(c) supplementary qualitative analyses such as expert judgements on the shape of the
resulting obligor distribution, minimum obligor numbers per grade and avoidance of
undue concentration in the different grades or pools.
(d) institutions should describe and store the calibration sample associated with each
calibration segment. The calibration sample should be comparable to the current
portfolio in terms of obligor and transaction characteristics but should reflect at the same
time the likely range of variability as referred to in section 5.4 in order to ensure
compliance with Article 180(1) point (a) of Regulation (EU) No 575/2013.
54
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
81. Institutions should conduct the calibration before the application of MoC or PD-Floors.
82. For the purpose of determining the PD estimates by obligor grade or pool as referred to in
Article 180(1) point (a) and (2) point (a) of Regulation (EU) No 575/2013
the long-run average default rates estimated according to Chapter 4 should be used as
calibration target for each grade and pool in each calibration segment;
Explanatory box for consultation purposes:
Example for multiple calibration segments
An example for a calibration segment that doesn’t coincide with the range of application of a
rating system is briefly given in the following:
Consider a scoring rating system for large and international companies. This rating system
aggregates is the combination of a quantitative scorecard and a qualitative scorecard. Whereas
the data for the quantitative scorecard is coming from the balance of the company, the input for
the qualitative data is based on the analyst’s assessment of the company in question. The
aggregated scores of these scorecards is finally calibrated to a PD applying the function
𝑃𝑃𝑃𝑃(Score) =
1
1+𝑒𝑒 −(𝛼𝛼∗Score+β)
for model parameters 𝛼𝛼 and 𝛽𝛽.
Whereas these quantitative and qualitative risk drivers may apply universally for all institutions in
the world, it could be argued that companies located in emerging markets could have a different
level of default rates in comparison with companies in Europe or North America.
Following this argumentation, it is in principle possible to apply a different calibration (in this
example: different values for 𝛼𝛼 and 𝛽𝛽) to companies in emerging markets. This would result in
two different calibration segments: “Companies from Emerging Markets” and “Companies from
Developed Markets”. Clearly these two calibration segments form a partition of the range of
application of the rating system.
83. Where institutions derive PD estimates from realised losses and appropriate estimates of
LGDs in accordance with Article 161(2) and 180(2)(b) of Regulation (EU) No 575/2013 they
should use a RDS that includes realised losses on all defaults identified during the historical
observation period specified in accordance with section 6.4.1 and relevant drivers of loss.
84. In order to use direct PD estimates for the calculation of capital requirements in accordance
with Article 169(3) of Regulation (EU) No 575/2013, institutions should demonstrate that the
theoretical assumptions of the probability model underlying the estimation methodology are
satisfied to a sufficient extent in practice.
85. When using the approach of using direct PD estimates for the calculation of capital
requirements in accordance with Article 169(3) of Regulation (EU) No 575/2013, institutions
may apply either of the following methods:
55
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(a) calculate the long-run averages of one-year default rates required in Article 180(1) point
(a), (2) point (a) of Regulation (EU) No 575/2013) at a level other than obligor grade that
is appropriate for the application of the probability model;
(b) instead of explicitly calculating default rates, they aggregate all relevant default and nondefault information implicitly for the estimation of a model whose outcomes can be
proven to be obligor PDs with sufficient certainty.
86. Whichever of the methods referred to in paragraph 85 an institution uses, all requirements
for the long-run averages of one-year default rates should then apply to the long-run
averages of one-year default rates calculated explicitly at the respective level, or, mutatis
mutandis, to the implicit incorporation of long-run one-year default information in the model
estimation. In particular, all data and representativeness requirements, including those in
accordance with Article 174 point (c) of Regulation (EU) No 575/2013, have to be met, and
the default definition in accordance with Article 178 of Regulation (EU) No 575/2013 has to be
used at the level at which the long-run one-year default information is incorporated for PD
estimation purposes. Under no circumstances can the use of continuous PDs or any default
rates smoothening be performed in order to overcome lack of data, low discriminatory
capacity or any other deficiencies in the rating or PD estimation process, or in order to reduce
the capital requirements.
87. Where institutions use segmentation drivers in the calibration process that define subsets of
the range of application of a given rating model where the same ranking model is applied as
in other subsets but each subset carries a significantly different level of risk from them, all of
the following should apply:
(a) more than one calibration in a given rating system should be used, where appropriate;
(b) the range of application of a PD model should be partitioned by as many calibration
segments as needed to obtain homogeneous calibration segments;
(c) the model should be calibrated separately for each calibration segment;
(d) the use of several calibration segments should be well documented.
88. Where scoring models are used, institutions should ensure that:
(a) where there is a change in the models used, the institutions consider whether it is
necessary to recalculate scores of obligors or facilities based on the original data set
instead of using scores that were calculated based on previous versions of the rating
system, and, where this is not possible, that institutions assess potential effects and
consider those effects by an appropriate increase of the MOC to their PD estimates;
(b) in the course of model adjustment the rank ordering is done in the score assignment and
that the calibration does not change that rank order;
56
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(c) where Article 180(2) point (g) or Article 180(1) point (g) of Regulation (EU) No 575/2013
apply, that PD estimates are adequate for grades which were derived as a simple average
of individual PD estimates, by applying calibration tests to this estimates on the basis of
one-year-default rates representative of the likely range of variability.
57
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
6. LGD estimation
6.1 General requirements for LGD estimation
89. Institutions that have obtained permission to use own estimates of LGD in accordance with
Article 143(2) of Regulation (EU) No 575/2013 should assign an LGD estimate to each nondefaulted exposure and an estimate of LGD in-default and ELBE to each defaulted exposure
within the scope of the rating system subject to such permission. Institutions should estimate
LGDs for all facility grades of the distinct facility rating scale or for all pools that are
incorporated in the rating system. For the purpose of LGD estimation institutions should treat
each defaulted facility as a distinct default observation, unless more than one independent
defaults were recognised on a single facility that do not meet the conditions of paragraph 90.
90. For the purpose of LGD estimation institutions should consider an exposure that after the
return to non-defaulted status is classified as defaulted again as having been constantly
defaulted from the first moment when the default occurred if the time between the moment
of the return of the exposure to non-defaulted status and the subsequent classification as
default is shorter than 1 year in any case. Institutions may specify a longer period than one
year for the purpose of considering two subsequent defaults as one for the purpose of LGD
estimation, if this is adequate to the specific type of exposures and reflects the economic
meaning of the default experience.
91. Institutions should estimate their own LGDs based on their own loss and recovery experience
that is reflected in historical data on defaulted exposures. Institutions may supplement their
own historical data on defaulted exposures with external data. In particular, institutions
should not derive their LGD estimates only from the market prices of financial instruments,
including, but not limited to, marketable loans, bonds or credit default instruments, but they
may use this information to supplement their own historical data.
92. Where in the case of retail exposures and purchased corporate receivables institutions derive
LGD estimates from realised losses and appropriate estimates of PDs in accordance with
Articles 161(2) and 181(2)(a) of Regulation (EU) 575/2013 they should ensure that:
(a) the process for estimating total losses meet the requirements of Article 179 of Regulation
(EU) 575/2013 and the outcome is consistent with the concept of LGD as set out in Article
181(1)(a) of this Regulation as well as with the requirements specified in Chapter 6, in
particular with the concept of economic loss as specified in section 6.3;
(b) the process for estimating PD meets the requirements of Articles 179 and 180 of
Regulation (EU) 575/2013 as well as the requirements specified in Chapter 5.
58
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
6.2 Data requirements for LGD estimation
6.2.1
Reference Data Set
93. For the purpose of LGD estimation institutions should use a Reference Data Set (RDS) covering
all of the following items:
(a) all defaults identified during the historical observation period specified in accordance with
section 6.4.1;
(b) all necessary data for calculating realised LGDs in accordance with paragraphs 112 to126;
(c) relevant factors that can be used to group the defaulted exposures in meaningful ways,
including their values at the moment of default and at least within the year before default
when available;
(d) relevant drivers of loss, including their values at the moment of default and at least within
the year before default when available.
94. Institutions should include in the RDS information on the results of the recovery processes,
including recoveries and costs, related to each individual defaulted exposure. To this end
institutions should include:
(a) information on the results of incomplete recovery processes until the reference date for
the LGD estimation.
(b) information on the results of recovery processes at portfolio level, where such
aggregation of the information is justified in particular in the case of indirect costs and
sale of a portfolio of credit obligations. When aggregated information is collected and
stored, institutions should develop an appropriate methodology for the allocation of
recoveries and costs to individual defaulted exposures and should apply this methodology
consistently across exposures and over time, ensuring, to the satisfaction of the
competent authority, that the methodology does not lead to biased LGD estimates.
(c) Information on external or pooled data used in the estimation of LGDs.
95. The RDS should contain at least the following information:
(a) obligor-related, transaction-related and institution-related risk characteristics as well as
external factors as referred to in paragraph 142 that are potential risk drivers at the
relevant reference dates as specified in paragraph 143;
(b) moment (date) of default;
(c) all default triggers that have occurred, including past due events and unlikeness to pay
events, even after the identification of default; in the case of exposures subject to
59
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
distressed restructuring the amount by which the financial obligation has diminished
calculated in accordance with the EBA Guidelines on the definition of default;
(d) the outstanding amount of the exposure at the moment of default including principal,
interest and fees;
(e) the amounts and timing of interest and fees capitalised after the moment of default;
(f) the amounts and timing of the additional drawings after default;
(g) the amounts and timing of write-offs;
(h) the values of collaterals associated with the exposure and, where applicable, the type of
valuation (such as market value or mortgage lending value as defined in points (74) and
(76) of Article 4(1) of Regulation (EU) 575/2013), date of valuation, a flag whether the
collateral has been sold and the sale price;
(i) information on any dependence between the risk of the obligor and the risk of the
collateral or collateral provider;
(j) the types, amounts and maturities of unfunded credit protection including the
specification and credit quality of the protection provider;
(k) the amounts, timing and sources of recoveries;
(l) the amounts, timing and sources of direct costs associated with recovery processes;
(m) a clear identification of the type of termination of the recovery process;
(n) where applicable, currency mismatches between two or more of the following elements:
the currency unit used by the institution for financial statements, the underlying
obligation and the funded or unfunded credit protection;
(o) amount of realised loss.
96. Institutions should collect and store in the RDS the information on the most recent evaluation
of the collateral before the moment of default except where the most recent evaluation was
performed shortly before default and was triggered by the decreased credit quality of the
obligation. In that case the previous evaluation from before the decrease of credit quality of
the obligation should be collected and stored in the RDS and used for the purpose of LGD
estimation in accordance with section 6.6.2.
97. In accordance with Article 229(1) of Regulation (EU) 575/2013 institutions may use various
methods for the valuation of the collateral in the form of immovable property including in
particular market value or mortgage lending value as defined in points (74) and (76) of Article
4(1) of that Regulation. Where institutions use various valuation approaches with regard to
60
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
immovable properties that secure exposures included in the scope of application of the
relevant rating system, they should collect and store in the RDS the information on the type
of valuation and use this information consistently in the LGD estimation and in the use of LGD
estimates.
98. Where institutions derive LGD estimates from realised losses and appropriate estimates of
PDs in accordance with Articles 161(2) and 181(2)(a) of Regulation (EU) 575/2013 they should
use a RDS that includes realised losses on all defaults identified during the historical
observation period specified in accordance with section 6.4.1 and relevant drivers of loss.
6.2.2
Representativeness of data
99. Institutions should perform an appropriate analysis to ensure that the data used for the
purpose of LGD estimation is sufficiently representative to the current portfolio covered by
the relevant LGD model. Such analysis should be performed separately for internal and
external data as well as for data from different sources.
100.
Institutions should analyse the representativeness of the data referred to in Article
179(1)(d) of Regulation (EU) No 575/2013 in all of the following manners:
(a) in terms of the scope of application;
(b) in terms of the definition of default;
(c) in terms of the distribution of the relevant risk drivers;
(d) in terms of lending standards and recovery policies;
(e) in terms of current and foreseeable economic or market conditions.
101.
For the purpose of paragraph 100(a) institutions should analyse the segmentation of
exposures into rating systems and should in particular consider whether there were any
changes to the scope of application of the LGD model over the historical observation period.
Where such changes were observed institutions should compare the distribution of the
characteristics of the relevant risk drivers in the RDS before and after the change.
102.
For the purpose of paragraph 100(b) institutions should analyse the definition of default
applied by the institution to the rating system currently and over the historical observation
period. In the case the definition of default has changed during the historical observation
period institutions should assess the representativeness of historical data included in the RDS
in the same way as specified for external data in Chapter 6 of EBA Guidelines on the
application of the definition of default under Article 178 of Regulation (EU) No 575/2013. In
the case the definition of default has changed during the historical observation period more
than once institutions should perform the analysis of each of the past definitions of default
separately.
61
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
103.
For the purpose of paragraph 100(c) institutions should analyse the distribution of the
characteristics of the relevant risk drivers in the RDS in comparison to the distribution of such
characteristics in the current portfolio of exposures that are within the scope of the rating
system. Institutions should perform such analysis separately for non-defaulted and defaulted
exposures in accordance with the following principles:
(a) In analysing representativeness of the RDS to the current portfolio of non-defaulted
exposures institutions should take into account the current characteristics of the risk
drivers at the moment of LGD estimation;
(b) In analysing representativeness of the RDS to the current portfolio of defaulted exposures
institutions should take into account the characteristics of the risk drivers at the relevant
reference date for each risk driver as specified in paragraph 143.
104.
Institutions should specify in their internal policies the statistical tests and metrics that
are used for the purpose of assessment of the representativeness of the data in terms of the
structure of the portfolio by relevant risk drivers. Where the application of statistical tests is
not possible institutions should at least carry out a qualitative analysis on the basis of the
descriptive statistics of the structure of the portfolio and present it in a graphic form. In the
analysis institutions should take into account possible seasoning effects of the relevant risk
drivers. When considering the results of the analysis institutions should take into account the
sensitivity of the risk drivers to economic conditions and the composition of portfolios of
defaulted and non-defaulted exposures.
105.
For the purpose of paragraph 100(d) institutions should analyse whether there were
significant changes in the lending standards or recovery policies over the historical
observation period that may influence the average loss rates or the distribution of the
characteristics of relevant risk drivers in the portfolio covered by the rating system. Where
institutions observe such changes they should compare the data included in the RDS before
and after the change of the policy with regard to average duration of the recovery processes,
frequency of use of certain recovery scenarios as well the loss severity distribution.
106.
Where during the historical observation period there were significant changes in the
relevant legal environment, such as changes in bankruptcy law or legal foreclosure
procedures, institutions should perform similar analysis to the one described in paragraph
105.
107.
Where external or pooled data are used institutions should obtain from the vendors
sufficient information about the lending and recovery policies of the data contributors in
order to assess representativeness of such external or pooled data to the institutions’ own
portfolios and processes.
108.
For the purpose of paragraph 100(e) institutions should consider all economic and market
conditions experienced in the past and reflected in historical observations as part of
foreseeable economic and market conditions.
62
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
109.
In the analysis of the representativeness of data institutions should take into account not
only the current characteristics of the portfolio but also, where relevant, the changes to the
structure of the portfolio that are expected to happen in the foreseeable future due to
specific actions or decisions that have already been taken. Adjustments made on the basis on
the changes expected to happen in the foreseeable future should not lead to decrease in LGD
estimates.
110.
Institutions should document the performed analysis and the measures taken to ensure
representativeness of the estimation sample to the current portfolio with respect to the
elements listed in paragraph 100 as well as changes to the respective practices. This
documentation should include the criteria and triggers established by the institution in order
to decide whether the data is sufficiently representative.
111.
In accordance with Article 181(1)(a) of Regulation (EU) 575/2013 institutions are required
to use all defaults observed in the historical observation period for the purpose of LGD
estimation that fall within the scope of the LGD model. Where historical data is not
sufficiently representative of a current portfolio institutions should provide, to the extent
possible, appropriate adjustments. In addition to these appropriate adjustments institutions
should increase the margin of conservatism applied to their LGD estimates in accordance with
Article 179(1)(a) and (f) of Regulation (EU) No 575/2013 as further specified in section 4.4.
Explanatory box for consultation purposes
The representativeness of historical observations to the current portfolio and situation can be
considered from various dimensions and the main dimensions have been specified in the draft
Guidelines. As non-representativeness in any of these dimensions may lead to a bias in LGD
estimates these have to be duly analysed and addressed. As in accordance with Article 181(1)(a)
of Regulation (EU) 575/2013 institutions are required to use all observed defaults it is not possible
to remove the observations that are not fully representative from the estimation sample.
However, in this case institutions should apply adequate margin of conservatism to account for
the weaknesses in data and, if possible, adjust the data to ensure greater representativeness.
The historical observations that are the basis for estimation of risk parameters should be
sufficiently representative to the whole portfolio that the estimates will apply to. Where different
portfolios are merged under one rating system institutions should analyse whether the rating
system is adequate for all parts of the portfolio that is within the scope of rating system. Where
the sample of historical observations is extended by additional data from a different source the
representativeness of these data from different sources should be assessed separately and where
necessary adequately adjusted.
Question 6.1: Do you agree with the proposed principles for the assessment of the
representativeness of data?
63
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
6.3 Calculation of economic loss and realised LGD
6.3.1
Definition of economic loss and realised LGD
112.
For the purpose of LGD estimation as referred to in Article 181(1)(a) of Regulation (EU) No
575/2013, institutions should calculate realised LGDs for each exposure, as referred to in
point (55) of Article 4(1) of that Regulation, as a ratio of the economic loss to the outstanding
amount of the credit obligation at the moment of default, including any amount of principal,
interest or fee.
113.
For the purpose of paragraph 112, institutions should calculate the economic loss realised
on an instrument (i.e. defaulted facility), as referred to in point (2) of Article 5 of Regulation
(EU) No 575/2013 as a difference between, from one side, the outstanding amount of the
credit obligation at the moment of default, without prejudice to paragraph 119, including any
amount of principal, interest or fee, increased by material direct and indirect costs associated
with collecting on that instrument discounted to the moment of default and, from the other
side, any recoveries realised after the moment of default discounted to the moment of
default.
114.
Where, relating to a default event, any part of exposure has been forgiven or written off
before or at the date of default and the amount forgiven or written off is not included in the
outstanding obligation at the moment of default the amount of the exposure that was
forgiven or written off should be added to the outstanding obligation at the moment of
default included in the denominator of the realised LGD.
115.
Without prejudice to paragraph 90, in the case of exposures that return to non-defaulted
status institutions should calculate economic loss as for all other defaulted exposures with the
only difference that additional recovery cash flow is added to the calculation at the date of
the return to non-defaulted status in the amount that was outstanding at the date of the
return to non-defaulted status. This additional recovery cash should not be discounted.
6.3.2
Treatment of unpaid late fees, interest and additional drawings after
default
116.
For the purpose of letter (i) of Article 181(1) of Regulation (EU) No 575/2013, institutions
should correct the economic loss by including in its calculation any fees that have been
capitalised in the institution's income statement after the moment of default and any
recoveries realised thereof. Institutions should not correct the outstanding amount of the
credit obligation at the moment of default in the denominator of the realised LGD. Where the
fees are extended to the obligor in order to recover direct costs already incurred by the
institution and these costs are already included in the calculation of the economic loss,
institutions should not add these amounts to the economic loss again.
117.
Institutions should apply the same treatment as specified in paragraph 116 to any interest
capitalised in the institution's income statement after the moment of default. In case of
64
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
recovery of late interest or fees that have not been previously capitalised the moment of
recovery should be considered a moment of capitalisation.
118.
In accordance with Article 182(1)(c) of Regulation (EU) No 575/2013 institutions that
obtained permission to use own estimates of LGD and conversion are required to reflect the
possibility of additional drawings by the obligor up to and after the time of default in their
estimates of conversion factors. In the case of retail exposures, in accordance with Article
181(2)(b) and Article 182(3) of this Regulation institutions may reflect future drawings either
in their conversion factors or in their LGD estimates. These future drawings should be
understood as additional drawings by the obligor after the moment of default.
119.
Where institutions include additional drawings by the obligor after the moment of default
in their conversion factors they should calculate realised LGD as a ratio of the economic loss
to the outstanding amount of the credit obligation at the moment of default increased by the
value of additional drawings by the obligor after the moment of default.
120.
For retail exposures, where institutions do not include additional drawings by the obligor
after the moment of default in their conversion factors they should calculate realised LGD as a
ratio of the economic loss to the outstanding amount of the credit obligation at the moment
of default and they should not increase the denominator of the ratio by the value of
additional drawings by the obligor after the moment of default.
121.
In any case institutions should calculate the economic loss used in the numerator of the
realised LGD including the additional drawings after the moment of default and all realised
recoveries.
Explanatory box for consultation purposes
As the concept of economic loss as well as realised LGDs are the basis for LGD estimation it is
considered necessary to specify these aspects clearly in the guidelines. The proposals in that
regard are based on the definitions included in point (55) of Article 4(1) of Regulation (EU) No
575/2013 and Article 5(2) of that Regulation. In order to reflect correctly the level of loss these
measures should also include events that occur after the moment of default such as additional
drawings, fees or interest.
For this purpose it has been specified that in accordance with Article 182(1)(c) of Regulation (EU)
No 575/2013 ‘additional drawings’ should be understood as drawings by the obligor up to and
after the moment of default. Where however this Regulation refers to ‘future drawings’ this
means only drawings after the moment of default. Hence the choice for the institutions specified
for retail exposures to reflect future drawings either in conversion factors or LGD estimates only
refers to additional drawings after the moment of default while all additional drawings up to the
moment of default have to be reflected in the conversion factors.
In specifying the treatment of additional drawings or fees after the moment of default it was
considered that the resulting measure of realised LGD should be consistent with the exposure
65
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
value that will be used for the purpose of calculation of capital requirements. Hence, where
additional drawings after default are included in the conversion factors the outstanding amount
in the denominator of the realised LGD should also include such drawings. However, in the case of
retail exposures, where the conversion factors do not include any drawings after the moment of
default the denominator of the realised LGD should also reflect only the outstanding amount at
the moment of default. As a result, lower exposure value (based on lower conversion factors) will
be compensated by higher LGD.
Similarly in the case of additional fees or interests that are capitalised after the moment of
default, as these are generally not included in the conversion factors they should not increase the
outstanding amount at the moment of default in the denominator of realised LGD. It is proposed
that such fees or interests should be treated similarly as costs, i.e. they should increase the
measure of exposure and loss in the numerator of the realised LGD, just as all recoveries related
to those fees, adequately discounted, should be included in the calculation of economic loss.
In any case, regardless of the choice of the institution regarding the estimation of conversion
factors, the measure of economic loss should be an objective value that adequately reflects the
actual loss experienced by the institution. Hence, the measure of economic loss should always
include all additional drawings, interest and fees capitalised after the moment of default as well
as all recoveries realised on a given exposure.
An alternative approach that was taken into account in specifying the draft guidelines was not
increasing the economic loss by the amount of fees or interests after default. In this case any
recoveries of these amounts would compensate the discounting effect that reflects the value of
money in time. As a result similar cases characterised by similar cash flows would lead to equal
results in terms of the realised LGD regardless of the pricing policy of the institution. However,
due to differences between the discounting rate and the interest rate applicable after default the
value of money in time and the recoveries related the amount outstanding at the moment of
default would not be reflected correctly. As the amount of additional interests calculated after
default would be discounted with a different discounting rate this would result in either a positive
or a negative correction of loss at the moment of default. In some cases this approach would lead
to negative realised LGD even if the outstanding amount was not fully recovered. Therefore this
approach was deemed not sufficiently prudent.
Question 6.2: Do you agree with the proposed treatment of additional drawings after default
and interest and fees capitalised after the moment of default in the calculation of realised
LGDs?
6.3.3
Discounting rate
122.
For the purpose of the calculation of economic loss, in accordance with point (2) of Article
5 of Regulation (EU) No 575/2013, institutions should discount all recoveries and costs,
including capitalised late fees and interest and additional drawings after the moment of
default using an annual discounting rate composed of a primary interbank offered rate
66
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
applicable at the moment of default increased by [5%-points] add-on. For this purpose the
primary interbank offered rate should be considered the 1-year EURIBOR or a comparable
interest rate in a currency of the exposure.
Explanatory box for consultation purposes
Discounting rate has been identified as one of the main drivers of non-risk based variability of the
LGD estimates. As various practices in that regard are observed several possibilities were
considered when specifying the draft Guidelines. The potential solutions taken into account
included in particular the following:
- original effective interest rate – despite the advantage of the consistency with international
accounting standards the use of the interest from the moment of origination of the loan was
considered not appropriate as this interest rate might not be adequate to the market and
economic conditions at the moment of default;
- current effective interest rate – although this interest is adequate to the moment of default it
still leads to lack of comparability of losses within and across institutions as the interest may
depend on various institutions pricing strategies and the rates may vary significantly between the
types of obligor and types of products; In addition to that in both cases the specification of
discounting rate based on effective interest rate would introduce substantial complexity as it
would be specified individually for each exposure;
- funding cost + add-on – the discounting rate reflecting the funding costs of the institution and an
appropriate risk premium reflecting the uncertainties associated with the receipt of recoveries
with respect to a defaulted exposure would be more consistent with the currently existing
guidance expressed by CEBS and BCBS however it was considered that it should not reflect the
own credit standing of the institution as this would lead to non-comparability across institutions
and jurisdictions; in addition, the discounting factor would be dependent on the funding structure
of a given institution;
- risk-free rate + add-on – this approach is independent from the funding structure and credit
standing of the institution and is in line with the currently existing guidance on the application of
the discounting factor that has been specified in GL10 and BCBS guidelines; However, the
specification of the add-on could still lead to significant variability across institutions and
jurisdictions.
Taking into account the above considerations it is proposed that the discounting rate should
follow the formula of risk-free rate + ad-on in accordance with the current guidance in that
regard. However, in order to ensure simplicity and comparability of the approach it is proposed
that the risk-free rate should be based on EURIBOR (or a comparable rate in non-Eurozone
countries) as a proxy of risk-free rate and the add-on should be a specified fixed percentage value
that reflects the average level of a risk premium.
67
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Although more granular approaches were considered for simplicity reasons the proposed fixed
value of the add-on is not differentiated between the exposure classes. In principle, it could be
possible to specify the discounting rate more granularly by implementing different predefined
add-on categories. Possible dimensions of this add-on slotting could include not only exposure
classes but also collateral types or types of products addressing different risk of the workout
processes. It was however deemed not practical due to different categorization of collaterals and
products across institutions.
Question 6.3: Do you agree with the proposed specification of discounting rate? Do you agree
with the proposed level of the add-on over risk-free rate? Do you think that the value of the
add-on could be differentiated by predefined categories? If so, which categories would you
suggest?
6.3.4
Direct and indirect costs
123.
For the purpose of the calculation of the realised LGDs, institutions should take into
account all material direct and indirect cost related to the recovery process. In the case any
material direct or indirect costs relating to the collection on exposures and the default of the
respective counterparty have been incurred before the moment of default institutions should
include these costs in the LGD estimation unless at least one of the following conditions is
met:
(a) these costs are clearly included in the exposure value;
(b) these costs are associated with the previous default of the same obligor that is not
considered a multiple default in accordance with paragraph 90.
124.
Direct costs should include the costs of outsourced collection services, legal costs, the
cost of hedges and insurances and all other costs directly attributable to the collection on a
specific exposure. Institutions should consider all direct costs as material.
125.
Indirect costs should include all costs stemming from the running of the institution’s
recovery processes, overall costs of outsourced collection services, and all other costs related
to the collection on defaulted exposures that cannot be directly attributed to collection on a
specific exposure. Institutions should include in their estimation of indirect costs an
appropriate percentage of other ongoing costs such as institution’s overheads related to the
recovery processes, unless they can demonstrate that these costs are immaterial.
126.
Institutions should demonstrate that they collect and store in their databases all
information required to calculate direct and indirect costs. All material indirect costs should
be allocated to respective exposures. This cost allocation process should be based on the
same principles and techniques that institutions use in their own cost accounting systems.
127.
For the purpose of indirect cost allocation institutions may use methods based on
exposure weighted averages, or statistical methods based on a representative sample within
68
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
the population of defaulted obligors. In any case institutions should demonstrate that the
indirect costs allocation process is effective and that it does not lead to biased LGD estimates.
6.4 Long-run average LGD
6.4.1
Historical observation period
128.
The historical observation period should be as broad as possible and should contain data
from various periods with differing economic circumstances. For this purpose institutions
should at a minimum select a historical observation period in such a way that:
(a) the length of the historical observation period, i.e. the timespan between the oldest
default considered in the RDS and the moment of the LGD estimation, covers at least the
minimum length specified in Article 181(1)(j) of Regulation (EU) No 575/2013 for
exposures to corporates, institutions, central governments and central banks and, for
retail exposure, the period specified in Article 181(2) subparagraph 2 of that Regulation
and, where applicable, Commission Delegated Regulation adopting technical standards
laid down in Article 181(3)(b) of that Regulation;
(b) it ensures that the estimation sample includes a sufficient number of closed recovery
processes in order to provide robust LGD estimates;
(c) it is composed of consecutive periods and includes the most recent periods before the
moment of LGD estimation;
(d) all available internal data is considered ‘relevant’, as referred to in Articles 181(1)(j) and
181(2) subparagraph 2 of Regulation (EU) No 575/2013 and is included in the historical
observation period.
129.
In assessing whether the estimation sample includes a sufficient number of closed
recovery processes in accordance with paragraph 128(b) institutions should take into account
both the absolute number of closed recovery processes as well as relative share of closed
recovery processes in the total number of observations.
Explanatory box for consultation purposes
The specification of the historical observation period is based on the assumption that it should be
as broad as possible and should contain data from various periods with differing economic
circumstances. These differing economic circumstances refer not only to the moment of default
but also to the moment of realising recoveries from different sources. In this context it was
considered that elimination of any data that reflects an institution’s internal experience would
lead to a loss of valuable information and hence it was specified that all available internal data
should be taken into account in the long-run average LGD.
69
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
An alternative approach to the historical observation period that was considered in developing
these Guidelines is on economic indicators, including either macroeconomic or credit factors. In
this approach the observation period would be specified in such a way that it would reflect the
likely range of variability of loss rates. Such observation period could be either longer or shorter
that the available data series. In the case the available data series do not cover the full range of
variability of loss rates institutions would have to perform appropriate adjustments to the longrun average LGD.
Eventually the simpler approach was chosen as more appropriate as in the estimation of LGD the
sufficiently broad sample of data is considered more important than the exact specification of the
historical observation period that reflects an economic cycle. This is because institutions are in
any case required to estimate LGDs that reflect economic downturn conditions if this is more
conservative than long-run average.
Question 6.4: Do you agree with the proposed approach with regard to the specification of
historical observation period for LGD estimation?
6.4.2
Calculation of long-run average LGD
130.
In accordance with letter (a) of Article 181(1) of Regulation (EU) No 575/2013 institutions
are required to calculate the long-run average LGD separately for each facility grade or pool.
In this context institutions should calculate the long-run average LGD also at the level of
portfolio covered by the LGD model.
131.
Without prejudice to Article 181(2) of Regulation (EU) No 575/2013 institutions should
calculate the long-run average LGD as an arithmetic average of realised LGDs over an
historical observation period weighted by a number of defaults. Institutions should not use
for that purpose any averages of LGDs calculated on a subset of observations, in particular
any yearly average LGDs, unless they use this method to reflect higher weights of more recent
data on retail exposures in accordance with Article 181(2) of Regulation (EU) No 575/2013.
132.
Where institutions do not give equal importance to all historical data for retail exposures
in accordance with Article 181(2) of Regulation (EU) No 575/2013 they should be able to
demonstrate in a documented manner that the use of higher weights to more recent data is
justified by better prediction of loss rates. In particular where a zero or very small weights are
applied to specific periods this should be duly justified or lead to more conservative
estimates.
133.
In specifying the weights in accordance with paragraph 132 institutions should take into
account the representativeness of data assessed in accordance with section 6.2.2 as well as
the economic and market conditions that are represented by the data.
70
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
6.4.3
Treatment of incomplete recovery processes
134.
For the purposes of letter (a) of Article 181(1) of Regulation (EU) No 575/2013, in relation
to the use of all defaults observed during the historical observation period within the data
sources for LGD estimation, institutions should ensure that the relevant information from
incomplete recovery processes is taken into account in a conservative manner. The LGD
estimation should be based on the long-run average LGD.
135.
Institutions should calculate the observed average LGD for each facility grade or pool and
at the level of portfolio covered by the LGD model taking into account realised LGDs on all
defaults observed in the historical observation period related to closed recovery processes in
accordance with paragraphs 136 and 137 without including any expected future recoveries.
The observed average LGD should be weighted by the number of defaults included in the
calculation.
136.
Institutions should clearly specify in their internal policies the moment of closing the
recovery processes. All recovery process that have been closed should be treated as such for
the purpose of the calculation of the observed average LGD. The observations where the
institution does not expect to take any further recovery actions should be recognised as
closed recovery processes without undue delay.
137.
Institutions should define the maximum period of the recovery process for a given type of
exposures from the moment of default that reflects the expected period of time observed on
the closed recovery processes during which the institution realises the most of the recoveries,
without taking into account the outlier observations with significantly longer recovery
processes. The maximum period of the recovery processes should be specified in such a way
that ensures sufficient data for the estimation of the recoveries within this period for the
incomplete recovery processes. The length of the maximum period of the recovery processes
may be different for different types of exposures. This specification of the maximum period of
the recovery process should be clearly documented and supported by evidence of the
observed recovery patterns, and should be coherent with the nature of the transactions and
the type of exposures. All exposures that remain in defaulted status for a period of time
longer than the maximum period of the recovery process specified for this type of exposures
should be treated as closed recovery process for the purpose of calculation of the observed
average LGD, considering only the recoveries realised so far.
138.
Institutions should obtain the long-run average LGD by adjusting the observed average
LGD taking into account the information related to incomplete recovery processes and the
estimated future costs and recoveries on these exposures in accordance with the following
conditions:
(a) where the time from the moment of default until the moment of estimation is longer
than the maximum period of the recovery process specified for this type of exposures
institutions:
71
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
i.
should take into account all actually observed recoveries realized before or after
the maximum period of the recovery process;
ii.
should not estimate any future recoveries.
(b) where the time from the moment of default until the moment of estimation is shorter
than the maximum period of the recovery process specified for this type of exposures
they may estimate future recoveries both those stemming from the realisation of the
existing collaterals and those to be realised without the use of collaterals;
(c) for the purpose of estimation of the future costs and recoveries institutions should
analyse the costs and recoveries realised on these exposures until the moment of
estimation in comparison to the average costs and recoveries realised during similar
period of time on similar exposures; for this purpose institutions should analyse the
recovery patterns observed on both closed and incomplete recovery processes taking into
account only factually observed costs and recoveries;
(d) the assumptions that underlying the expected future costs and recoveries as well as the
adjustment to the observed average LGD should be:
i.
proven accurate through backtesting;
ii.
based on a reasonable economic rationale;
iii.
proportionate, taking into consideration that LGD estimates should be based on
the long-run average LGD that reflects the average LGDs weighted by the number
of defaults using all defaults observed during an historical observation period.
(e) in estimating the future recoveries institutions should take into account the potential bias
stemming from incomplete recovery processes being characterised by longer average
recovery processes and lower average recoveries in comparison to closed recovery
processes;
(f) in estimating the future recoveries stemming from the realisation of the existing
collaterals institutions should take into account the legal certainty of the collateral and
realistic assumptions regarding the possibility of its realisation;
(g) the adjustment of observed average LGD may be estimated at the level of individual
exposure, at the level of grade or pool or at the level of portfolio covered by the LGD
model;
(h) any uncertainty related to the estimation of the future recoveries on incomplete recovery
processes should be reflected in appropriate MoC applied in accordance with section 4.4.
72
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Explanatory box for consultation purposes
The specification of the treatment of incomplete recovery processes takes into account the
requirement of Article 181(1)(a) of Regulation (EU) No 575/2013 that the long-run average LGD
has to take into account all observed defaults. Hence, the long-run average LGD should also
include incomplete recovery processes. However, in order to obtain a realistic value of long-run
average LGD these incomplete recovery processes should be included with future recoveries that
are expected to be realised. The value of future recoveries has to be estimated based on the
observed closed cases. As a result, the ‘long-run average LGD’ will be a measure that is not fully
objective as it is already partly estimated.
In order to obtain a fully objective measure it is proposed that institutions should also calculate
the ‘observed average LGD’ taking into account realised LGDs only on those defaults that are
related to closed recovery processes. Although this objective measure will not include any
elements of estimation it has to be kept in mind that it may not reflect the real experience
correctly as the cure and high-recovery cases may be overrepresented. More difficult cases
usually stay longer in recovery processes therefore they will more likely not be included in the
‘observed average LGD’.
Therefore the ‘observed average LGD’ has to be adjusted to account for the most recent
experience based on the incomplete recovery processes. It may also be noted that where the
share of incomplete recovery processes in the sample is higher the measure of ‘observed average
LGD’ may be less reliable.
Question 6.5: Do you agree with the proposed treatment of incomplete recovery processes in
obtaining the long-run average LGD?
6.4.4
Treatment of cases with no loss or positive outcome
139.
Where institutions observe that they realised profit on their observations of defaults the
realised LGD on these observations should equal 0 for the purpose of calculation of observed
average LGD and estimation of long-run average LGD.
6.5 LGD estimation methodologies and risk drivers
140.
Institutions should be able to demonstrate that the methodologies that they choose for
the purpose of LGD estimation are appropriate to their activities and the type of exposures to
which the estimates apply and they should be able to justify the theoretical assumptions of
those methodologies. The LGD estimation methodologies should in particular be consistent
with the collection and recovery policies adopted by the institution and should take into
account possible recovery scenarios as well as potential differences in the legal environment
in relevant jurisdictions.
141.
The functional and structural form of the estimation method used by the institution in the
LGD estimation, the assumptions regarding this method, its downturn effect, the length of
73
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
data series used, the margin of conservatism, the human judgement and, where applicable,
the choice of risk drivers, should be adequate to the type of exposures to which they are
applied.
142.
Institutions should identify and analyse potential risk drivers that are relevant to its
specific circumstances and to the specific characteristics of the type of exposures covered by
the rating system. Potential risk drivers analysed by institutions should include in particular
the following:
(a) Transaction-related risk characteristics, including type of product, type of collateral,
geographical location of the collateral, unfunded credit protection, seniority, Loan-toValue ratio (LTV), exposure size, seasoning, and recovery procedures;
(b) Obligor-related risk characteristics, including, where applicable, size, capital structure,
geographic region, industrial sector, and line of business;
(c) Institution-related factors, including internal organisation and internal governance,
relevant events such as mergers, and existence of specific entities within the group
dedicated to recoveries such as ‘bad credit institutions’;
(d) External factors, including interest rates, legal framework and other factors influencing
expected length of the recovery process;
143.
Institutions should analyse the risk drivers not only at the moment of default but also at
least within a year before default. Institutions should use a reference date for a risk driver
that is representative of the realisations of the risk driver within a year before default. When
choosing the appropriate reference date for a risk driver institutions should take into account
its volatility over time.
144.
Institutions should specify or calculate the risk drivers in the same way in the application
of LGD estimates as they are specified or calculated in the estimation of LGD.
6.6 Treatment of collaterals in LGD estimation
6.6.1
Eligibility of collaterals
145.
In accordance with Article 181(1)(f) of Regulation (EU) No 575/2013 institutions may take
into account in their LGD estimations the existence of any types of collaterals for which they
have established internal requirements in terms of collateral management, legal certainty and
risk management that are generally consistent with those set out in Section 3 of Chapter 4 of
Title II in Part Three of that Regulation. In the case of the types of collateral that are not
specified in Chapter 4 of Title II in Part Three of that Regulation institutions may use them in
their LGD estimations where their policies and procedures relating to internal requirements
for valuation and legal certainty of these collaterals are appropriate to this type of collateral.
74
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
146.
To the extent the LGD estimates take into account the existence of unfunded credit
protection institutions should specify the criteria and methodology for recognising and
including in their LGD estimates the protection in the form of guarantees and credit
derivatives that meet the criteria specified in Article 60 of Commission Delegated Regulation
xxx/xxxx [RTS on IRB assessment methodology].
6.6.2
Inclusion of collaterals in the LGD estimation
147.
Institutions should take into account as a risk driver or segmentation criterion information
on all main types of collaterals that are used within the scope of application of the LGD
model. Institutions should clearly define in their internal policies the main and other types of
collaterals used for the type of exposures covered by the rating system and should ensure
that, to the extent that LGD estimates take into account the existence of collateral, the
policies regarding the management of these types of collateral comply with the requirement
of Article 181(1)(f) of Regulation (EU) No 575/2013. Institutions should specify the main types
of collaterals in such a way that the cash flows from the remaining types of collaterals will not
significantly bias the estimation of recoveries that are realised without the use of collaterals.
148.
For the purpose of LGD estimation institutions may group the types of collaterals that are
homogeneous in terms of recovery patterns taking into account both the average time of
collection process and the recovery rates on these types of collaterals.
149.
The approach developed by the institutions to include the effect of collaterals in the LGD
estimation should meet all of the following conditions:
(a) Institutions should avoid the bias that may stem from including the cash flows related to
realisation of collateral in the estimation of recoveries that are realised without the use of
collaterals and the other way round.
(b) In estimating the recovery rates related to specific types of collaterals institutions should
avoid a bias that may stem from including in the estimation sample the observations
where the exposure was secured by only a part of the value of the collateral; in these
cases the estimation should be based on the total value of the collateral and total sale
price of the collateral.
(c) Where institutions estimate separately recovery rates related to specific types of
collaterals they should recognise and include in this estimation direct costs related to the
collection on these types of collaterals.
(d) Where institutions estimate separately recovery rates related to specific types of
collaterals they should include in this estimation all recoveries realised from a respective
type of collateral including those realised on exposures where the realisation of the
collateral has been completed but the overall recovery process in not yet closed.
75
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(e) The estimates should not be based solely on the estimated market value of the collateral
but they should also take into account the realised recoveries from past liquidations and
the potential inability of an institution to gain control and liquidate the collateral. In
estimating the recovery rates related to specific types of collaterals institutions should
take into account the time between the moment of default and the time of cash flows
related to the collection on these types of collaterals.
(f) The estimates should take into account the potential decreases in collateral value from
the point of LGD estimation to the eventual recovery, in particular those resulting from
the changes in the market conditions, state and age of the collateral and, where relevant,
currency fluctuations. The estimates should not take into account any potential increases
in collateral value from the point of LGD estimation to the eventual recovery.
(g) The estimates should take into account in a conservative manner the degree of
dependence between the risk of the obligor and collateral as well as the cost of
liquidating the collateral.
Explanatory box for consultation purposes
In accordance with Article 179(1)(a) of Regulation (EU) No 575/2013 institutions are required
to incorporate in their LGD estimates all relevant data, information and methods. As the
existence of collaterals significantly influences the level and timing of recoveries information
on at least the main types of collaterals used for a given type of exposures should be
considered relevant and included in the LGD estimates.
In addition to that, in accordance with Article 181(1)(f) of Regulation (EU) No 575/2013, to the
extent that LGD estimates take into account the existence of collateral, institutions shall
establish adequate internal requirements for collateral management, legal certainty and risk
management. It has been further clarified in the RTS on IRB assessment methodology that
these internal requirements should be at least fully consistent with the requirements of
Section 3 of Chapter 4 of Regulation (EU) No 575/2013 with regard to legal certainty and
regular valuation of collateral. Therefore, such adequate policies should be implemented for
all main types of collaterals as well as all other collaterals that are incorporated in the LGD
estimates.
Question 6.6: Do you agree with the proposed principles on the treatment of collaterals in
the LGD estimation?
6.6.3
Cash flows from collaterals
150.
Institutions should recognise the recoveries as stemming from collaterals in all of the
following situations:
(a) the collateral is sold by the obligor and the obtained price has been used to cover the
defaulted exposure;
76
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(b) the collateral is repossessed or sold by the institution, the parent undertaking or any of its
subsidiaries;
(c) the collateral is sold in the court or bailiff procedure;
(d) the credit obligation is sold and the price for the obligation included the existing collateral
– in this case it may be necessary to specify an allocation methodology in order to
determine which part of the price received for the sold obligations was related to the
existing collateral;
(e) any other method of realising the collateral possible in the legal framework in a relevant
jurisdiction.
151.
For the purpose of point (b) of paragraph 150 institutions should consider the value of
repossession the value by which the credit obligation of the obligor has been diminished as a
result of the repossession of the collateral, and which the repossessed collateral was recorded
as an asset on the balance sheet of the institution. Where these values are different
institutions should consider as a value of repossession the lower of them. The value of
repossession should be considered a value of recovery at the date of repossession and should
be included in the calculation of the economic loss and realised LGD in accordance with
section 6.3.
152.
Institutions should consider whether the value of repossession adequately reflects the
value of the repossessed collateral, consistently with any established internal requirements
for collateral management, legal certainty and risk management. In the case there is
significant uncertainty whether the value of repossession adequately reflects the value of the
repossessed collateral, institutions should apply an appropriate haircut to this value and
include in the calculation of economic loss a recovery as a value of repossession after the
haircut. Institutions should estimate this haircut taking into account all of the following
conditions:
(a) the haircut should reflect the possible errors in the valuation of the collateral at the
moment of repossession taking into account the type of the valuation available at the
moment of repossession, the date it was performed and the liquidity of the market for
this type of asset;
(b) the haircut should be estimated with the assumption that the institution intends to sell
the repossessed collateral to an independent third party and should reflect the potential
price that could be achieved from such sale, the costs of the sale and the discounting
effect to the moment of repossession taking into account the liquidity of the market for
this type of assets;
(c) where there are observations available regarding the repossessions and subsequent sales
of similar types of collaterals the estimation of the haircut should be based on these
77
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
observations and regularly backtested; for this purpose institutions should take into
account all of the following:
(i) difference between the value of repossession and the sale price;
(ii) any income and costs related to this asset that were observed between the date
of repossession and the moment of the sale;
(iii) discounting effects;
(iv) whether the institution repossessed the collateral with the intension of
immediate sale or whether another strategy was adopted.
(d) where the historical observations regarding the repossessions and subsequent sales of
similar types of collaterals are not available the estimation of the haircut should be based
on a case-by-case assessment, including the analysis of the current market and economic
conditions;
(e) the less data an institution has on the previous repossessions and the less liquid is the
market for the given type of assets the more uncertainty is attached to the resulting
estimates, which should be adequately reflected in MoC in accordance with section 4.4..
153.
In any case the repossession of collateral should be recognised at the moment of
repossession and should not prevent the institution from closing the recovery process in
accordance with paragraph 136.
154.
Institutions should make every effort to recognise the sources of the cash flows and
allocate them adequately to the specific collateral or unfunded credit protection that has
been realised. Institutions should specify clear policies for the treatment of cash flows where
the source cannot be identified. The methodology of allocation of these cash flows should not
lead to a bias in LGD estimation.
155.
For the purpose of calculation of the economic loss realised on an exposure in accordance
with paragraph 113 institutions should take into account all realised recoveries including the
recoveries from unknown sources and recoveries related to collaterals that do not meet the
requirement of Article 181(1)(f) of Regulation (EU) No 575/2013.
Explanatory box for consultation purposes
When developing these Guidelines the EBA considered whether the repossession of collateral by
an institution should be treated as a recovery. On the one hand at the moment of repossession
the credit obligation is decreased by the value of repossessed collateral and a new asset is
registered on the institution’s balance sheet that is separately risk-weighted. On the other hand
the institution has not realised any cash flows yet and additional loss may be realised at the
78
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
moment of the sale of this asset. Moreover, the time between the repossession and the final sale
of the repossessed collateral may be significantly long.
In order to address all of these considerations it is proposed that the repossession should be
treated as a recovery at the moment of repossession. As a result, the recovery processes may be
recognised as closed before the final sale of the asset that may take place much later. However,
where there is uncertainty about the true value of the repossessed collateral and there are
concerns that additional loss may be incurred as a result of the sale of this asset, institutions
should apply an adequate haircut to the value of recovery to reflect this uncertainty.
It is assumed under the approach taken that the lower of (i) the value by which the credit
obligation of the obligor has been diminished as a result of the repossession of the collateral and
(ii) the value of the repossessed collateral as recorded as an asset on the balance sheet of the
institution reflects the market value at the date of repossession, if necessary after application of
an appropriate haircut.
According to the proposed approach the value of repossession, subject to a haircut where
necessary, is treated as a recovery. The haircut does not capture the potential future fluctuations
of the prices for a certain type of assets. Hence, if the institution subsequently to the repossession
keeps the asset on its balance sheet the risk related to this asset is considered investment risk not
related to the original credit obligation of the obligor.
The alternative solution that was taken into consideration in the process of developing these
guidelines was that to measure the risk up to the sale of the asset. According to this approach the
haircut would reflect not only the uncertainty around the value of the asset at the moment of
repossession but also any potential future changes in market prices. Consequently the final
recovery would be considered at the moment of the sale of the asset and the default observation
in the RDS would have to be modified at this point in time to reflect the sale of the asset as a
recovery instead of the value of repossession. The modified recoveries would have to reflect the
sale price as well as any income and costs related to this asset that were observed between the
date of repossession and the moment of the sale.
Question 6.7: Do you agree with the proposed treatment of repossessions of collaterals? Do
you think that the value of recovery should be updated in the RDS after the final sale of the
repossessed collateral?
6.7 Downturn adjustment
156.
For the purpose of obtaining LGD estimates that are appropriate for an economic
downturn in accordance with Article 181(1)(b) of Regulation (EU) No 575/2013 and Article 52
of Commission Delegated Regulation xxx/xxxx [RTS on IRB assessment methodology]
institutions should specify an economic downturn in accordance with the RTS on the nature,
severity and duration of an economic downturn developed on the basis of Article 181(3)(a) of
Regulation (EU) No 575/2013.
79
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Explanatory box for consultation purposes
The EBA intends to publish the RTS on the basis of Article 181(3)(a) of Regulation (EU) No
575/2013 that will specify the common understanding of the nature, severity and duration of an
economic downturn. Institutions are required to estimate LGDs that are appropriate for such an
economic downturn. This could be performed, in particular, by specifying a downturn adjustment
to the long-run average LGD.
It could be considered whether the calibration of LGDs to downturn conditions should be based
on defaults identified in the downturn period or recoveries realised during this period. In the case
of the LGD methodology that is based on certain model components, whether downturn
adjustment should be estimated with relation to LGD or rather to an individual model
component? These choices have been left to institutions as different approaches may be
appropriate to different possible circumstances.
It has already been specified in Article 53(d) of the RTS on IRB assessment methodology that any
adverse dependencies between economic factors and recovery rates should be reflected in LGD
estimates. Hence, institutions should analyse such dependencies in various dimensions and with
various economic factors, including both macroeconomic and credit factors. These various
dimensions should include in particular total recoveries, recoveries realised from specific sources,
specific model components where appropriate and total losses, losses specified in terms of the
year of default and based on the time of recovery etc. The resulting downturn adjustment should
be specified appropriately depending on the level and character on the identified dependencies
with relevant economic factors.
Question 6.8: Do you think that additional guidance is necessary with regard to specification of
the downturn adjustment? If yes, what would be your proposed approach?
80
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
7. Estimation of risk parameters for
defaulted exposures
7.1 General requirements specific to ELBE and LGD in-default
estimation
157.
Institutions that have obtained permission to use own estimates of LGD in accordance
with Article 143(2) of Regulation (EU) No 575/2013, should assign an ELBE estimate and a LGD
in-default estimate to each defaulted exposure within the scope of the rating system subject
to such permission.
158.
Institutions should estimate ELBE and LGD in-default for each of the facility grades of the
distinct facility rating scale or for each of the pools that are incorporated in the rating system.
159.
For the purposes of ELBE and LGD in-default estimation, and unless otherwise specified in
this Chapter, institutions should use the same estimation methods used for estimating LGD on
non-defaulted exposures, as set out in Chapter 6.
Explanatory box for consultation purposes
It has already been specified in Article 54(2)(c) of the RTS on IRB assessment methodology that
the direct estimation of LGD in-default should be consistent with the methodology set for LGD for
non-defaulted exposures in order to avoid potential cliff effects. Following this approach, for the
purpose of both ELBE and LGD in-default estimation the Guidelines refer to the requirements on
LGD estimation of Chapter 6 including the same general estimation requirements of Chapter 4
and the requirements on the application of risk parameters of Chapter 8. This chapter should then
provide guidance only on those specific aspects where different treatment for defaulted assets is
justified. This implies, for example, that:

When estimating ELBE and LGD in-default, institutions should use their loss and recovery
experience (and may supplement this own experience with external data) and should not
derive their ELBE and LGD in-default from the market prices of financial instruments such as
marketable loans, bonds or credit default instruments;

When estimating ELBE and LGD in-default, institutions should use the same RDS and the same
methods for computing the long-run average LGD including, for example, the treatment of
incomplete recovery processes and the historical observation period.

LGD in-default estimates should include the appropriate margin of conservatism following the
same approach presented in section 4.4.
81
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Question 7.1: Do you agree with the proposed approach to the ELBE and LGD in-default
specification? Do you have any operational concerns with respect to these requirements? Do
you think there are any further specificities of ELBE and LGD in-default that are not covered in
this chapter?
160.
Institutions should take into consideration all relevant post-default information in their
ELBE and LGD in-default estimates in a timely manner. In particular, where events from the
recovery process invalidate the recovery expected by the most recent ELBE estimation,
institutions should update immediately their ELBE and LGD in-default estimates.
161.
Institutions should assess and duly justify situations where there are systematic
deviations of the LGD in-default estimates just after the date of default from the LGD
estimates just before the date of default at the facility grade or pool, which are not stemming
from the use of risk drivers that are applicable from the date of default onwards.
162.
Institutions should perform back-testing and benchmarking of their ELBE and LGD indefault estimates according to points (b) and (c) respectively, of Article 185 of Regulation (EU)
No 575/2013.
7.2 Data requirements specific to LGD in-default and ELBE
estimation
163.
For the purposes of ELBE and LGD in-default estimation, institutions should use the same
Reference Data Set (RDS) as referred to in section 6.2.1, complemented to reflect any
relevant information observed during the recovery process and at each reference date,
specified in accordance with paragraphs 164 to 167 and, in particular at least the following
additional information:
(a) all relevant factors that can be used to group defaulted exposures, including those that
may become relevant after the date of default and at each reference date;
(b) all relevant drivers of loss, including those that may become relevant after the date of
default and at each reference date;
(c) the amount outstanding at each reference date;
(d) the values of collateral associated with the exposures and their dates of valuation after
the date of default.
7.3 Reference dates
164.
For the purposes of ELBE and LGD in-default estimation, institutions should set the
reference dates that can be used to group defaulted exposures in a significant manner in
terms of the recovery pattern observed. For the purposes of setting the reference dates
institutions should use only closed recovery processes and those recovery processes that are
82
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
treated as closed in accordance with paragraph 137 and factually observed costs and
recoveries from incomplete recovery processes.
165.
Each of the reference dates referred to in paragraph 164 could be any of the following:
(a) a specific number of days since the date of default, which would be appropriate in case
the estimation refers to a portfolio of exposures that have a stable recovery pattern
through time and so reference date could be time in-default related;
(b) a relevant date associated with a specific event at which significant breaks in the recovery
profile are observed, which would be appropriate in case the estimation refers to a
portfolio of exposures that have significant changes of recovery pattern associated to
some specific events, for instance at the realisation of collateral;
(c) any combination of the cases referred to in points (a) and (b) that better reflects the
recovery pattern, which would be appropriate in case the estimation refers to a portfolio
of exposures that have a stable recovery pattern through time but on which a recovery
break-in is observed around certain specific events, for instance at collection, and where
the reference dates following those events are defined as specific number of days since
the recovery event rather than since the date of default;
(d) where appropriate, the reference date can have any value between zero and the
maximum number of days a recovery process on a defaulted exposure remains
incomplete.
166.
For the purposes of ELBE and LGD in-default estimation the same defaulted exposure in
the RDS should be used in all relevant grades or pools of exposures according to each
different reference date.
167.
Institutions should monitor on a regular basis potential changes in the recovery patterns
and in the relevant recovery policies which may affect the estimation of ELBE and LGD indefault at each reference date.
Explanatory box for consultation purposes
The difference between the LGD in-default and the ELBE (the unexpected loss) is used for
computing the risk weight which according to Article 153 (1)(ii) of Regulation (EU) No 575/2013
which applies to the current outstanding exposure amount (i.e. net of recovery from the date of
the default to the current time of evaluation). Moreover, the ELBE is compared to credit risk
adjustments for IRB shortfall purposes where credit risk adjustments are again computed with
respect to the current exposure amount. This forms the rationale for using a reference point for
the evaluation of the outstanding obligation that is close to the current time of evaluation rather
than the date of default for the purposes of computing realised losses used to estimate ELBE and
LGD in-default.
83
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
The concept of current outstanding exposure amount is clearly defined in Article 166(1) of
Regulation (EU) No 575/2013 and is used for the purposes of the application of the ELBE and LGD
in-default estimates. The Guidelines complement this by setting a definition of outstanding
exposure amount at each reference date for LGD and ELBE estimation purposes. Institutions
should set discrete relevant reference dates at which the realised LGDs could be computed on
each pool of exposures. These reference dates should be set according to the recovery patterns
observed on homogenous portfolios of exposures and will be used as basis for differentiating
grades or pools of exposures within this portfolio. This will imply that:
A.
For the purpose of estimating LGD in-default and ELBE: Institutions should calculate the
realised LGDs and related economic losses on all the defaulted exposures in the reference data
set and at each of the reference dates relevant for the pool to which the defaulted exposure
belong.
B.
For the purpose of application of the estimated LGD in-default and ELBE to a given
defaulted exposure in the current portfolio: Institutions should first evaluate which is the relevant
reference date for the exposure under consideration (e.g. if the exposure under consideration pre
or post collection from collateral). The ELBE (or LGD in-default) to be assigned to the exposure
under consideration should then be calculated as the product of the ELBE (or LGD in-default)
estimate at the relevant reference date (in % terms) and the current outstanding exposure value.
Similarly the risk weight to be applied to the current outstanding exposure is calculated according
to the difference between the LGD in-default and ELBE at the relevant reference date. Therefore it
is important that the same reference dates are used for the purpose of both the LGD in-default
and ELBE.
The expected loss on an exposure can change a lot from one reference date to the next, and
therefore the approach to setting the reference dates should fit the observed recovery patterns
and should be specified by institutions for specific types of exposures. Take as an example the
case of a portfolio of mortgage loans. The reference date for estimation will be most likely based
on the realisation of collateral event, such that to have an estimation of ELBE and LGD in-default
before and after the realisation of collateral, as paragraph 165(b) suggests. Otherwise it could be
based on a combination of reference dates set as function of time since default before the
realisation of collateral event and as a function of time since the realisation of collateral after the
realisation of the collateral. A reference date simply based on number of days in-default rather
than on the date of collateral realisation could in fact bias the ELBE estimates since the timing of
collateral realisation could be exposure specific. On the contrary take the example of a consumer
loan portfolio where all exposures are uncollateralised and characterised by a stable recovery
process through time. This recovery patterns justifies the use of reference dates defined as days
since default (e.g. quarterly frequency) as considered in paragraph 165.
Institutions should also monitor on a regular basis potential changes in the recovery patterns and
relevant policies (e.g. write off policies) which may affect the estimation of LGD in-default and
ELBE at each reference date.
84
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Question 7.2: Do you agree with the proposed reference date definition? Do you currently use
the reference date approach in your ELBE and LGD in-default estimation?
7.4 Calculation of realised LGD and long-run average LGD for
defaulted exposures
168.
For the purposes of ELBE and LGD in-default estimation institutions should calculate the
realised LGDs for defaulted exposures, in accordance with section 6.3 with the only difference
that this should be done with regards to the reference date, specified in accordance with
paragraphs 164 to 167, rather than the date of default.
169.
For the purposes of ELBE and LGD in-default estimation institutions should calculate the
long run average LGD of the realised LGDs for defaulted exposures, referred to in paragraph
168, following the requirements set out in section 6.4 with the only exception that
incomplete recovery processes should be used only for those reference dates beyond which
factual recovery and costs are observed.
Explanatory box for consultation purposes
It is proposed that for the purpose of estimating LGD in-default and ELBE institutions should
compute the realised LGD for defaulted exposures following the approach described in section 6.3
for non-defaulted exposures but with the difference that the calculation should be performed
with respect to the reference dates, as defined in paragraph 164 rather than the date of default.
This implies that institutions should calculate for each defaulted exposure the realised LGD
according to each reference date. This will be the ratio between the economic loss and the
amount of the credit obligation outstanding at the reference date where the economic loss is
computed as the difference between, on the one hand, the amount of the credit obligation
outstanding at the reference date (including any amount of principal, interests or fees) increased
by any direct and indirect costs associated with collecting on that instrument and discounted to
the reference date and, on the other hand, any recoveries realised after the reference date
discounted to the reference date. As a result for each defaulted exposure in the RDS institution
should compute as many realised LGDs as the number of relevant reference dates for estimation
purposes.
As specified in paragraph 169 in calculating the long run average LGD for defaulted exposures
institutions should follow the requirements specified in section 6.4 for non-defaulted exposures,
including the treatment of incomplete recovery processes (section 6.4.3). The only exception
envisaged in paragraph 169 with respect to the inclusion of incomplete recovery processes in the
ELBE and long run average LGD for defaulted exposures is that those can be included only with
respect to reference dates beyond which factual recovery and costs have been already observed.
This was put in place to avoid a circular reference of an estimation within the estimation. The
estimation of the future costs and recoveries on incomplete recovery processes should be
consistent between defaulted and non-defaulted exposures and should be based, as suggested in
paragraph 138(c), on a comparison of the costs and recoveries realised on these exposures until
85
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
the moment of estimation to the average costs and recoveries realised during similar period of
time on similar exposures. For this purpose institutions should analyse the recovery patterns
observed on both closed and incomplete recovery processes taking into account only observed
costs and recoveries.
An alternative approach to the treatment of incomplete recovery processes that was considered
in developing the Guidelines is that of using only closed recovery processes for the purposes of
estimating LGD in-default and ELBE. This approach has the advantage of simplicity in which it
avoids any estimation on future recoveries and costs for incomplete recovery processes. This
would come at the cost of disregarding the potential valuable information coming from
incomplete recovery processes and creating potential cliff effects between the LGD for defaulted
and non-defaulted exposures which are not due to additional information and risk drivers
available for non-defaulted exposures. Moreover, this approach may be less consistent with
Article 181(1)(a) of Regulation (EU) No 575/2013 where all observed default should be used for
the purposes of LGD estimation.
The specification of incomplete recovery processes of section 6.4.3 for non-defaulted exposures
takes into account the requirement of Article 181(1)(a) of Regulation (EU) No 575/2013. In
consideration to the fact that this Article should apply also for the purposes of LGD in-default and
ELBE estimation the approach of being consistent with the treatment of incomplete recovery
processes specified for non-defaulted exposures was chosen as more appropriate.
Question 7.3: Do you agree with the proposed approach with regard to the treatment of
incomplete recovery processes for the purpose of estimating LGD in-default and ELBE?
7.5 Risk drivers
170.
For the purposes of taking into account the information on the time in-default and
recoveries realised so far, in accordance with Article 54(2)(b) of Commission Delegated
Regulation xxx/xxxx [RTS on IRB assessment methodology] institutions may take into account
this information either directly as risk drivers or indirectly, for instance by setting the
reference date for estimation, as referred to in paragraphs 164 to 167.
171.
For the purpose of ELBE and LGD in-default estimation institutions should analyse the
potential risk drivers referred to in paragraph 142 not only until the moment of default but
also after the date of default and until the date of termination of the recovery process.
Institutions should analyse also other potential risk drivers that might become relevant after
the date of default, including in particular the expected length of the recovery process and
the status of the recovery process.
7.6 Specific requirements for ELBE estimation
172.
For the purpose of Article 181(1)(h) of Regulation (EU) No 575/2013 the ELBE should not
include any of the MoC referred to in section 4.4.
86
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
7.6.1
Current economic circumstances
173.
For the purposes of considering current economic circumstances in their ELBE estimates,
as required by Article 181(1)(h) of Regulation (EU) No 575/2013, institutions should take into
account economic factors, including macroeconomic and credit factors, relevant for the type
of exposures under consideration.
174.
Where the realised LGD for defaulted exposures, referred to in paragraph 168, is not
sensitive to the economic factors relevant for the type of exposures under consideration the
ELBE should be calculated on the basis of the long-run average LGD, as referred to in paragraph
169.
175.
Where the realised LGD for defaulted exposures, referred to into in paragraph 168, is
sensitive to the economic factors relevant for the type of exposures under consideration, the
institutions should adjust the long run average LGD for defaulted exposures such that to
reflect current economic circumstances. For this purposes institutions could use different
approaches, for example, including the relevant economic and credit factors for the exposure
under consideration in the estimation or considering risk drivers that are sensitive to those
economic and credit factors.
176.
Whichever of the approaches is used for the adjustment to current economic
circumstances institutions should document separately the long-run average LGD for
defaulted assets, referred to in paragraph 169, and the adjustment to current economic
circumstances, referred to in paragraph 175.
Explanatory box for consultation purposes:
Institutions are required to obtain ELBE estimates that are appropriate for current economic
circumstance. This should be performed, in particular, by specifying an adjustment to the long-run
average LGD for defaulted exposures.
Changes in economic conditions can be seen through changes in the values of risk drivers, e.g.
deviations of the observed loan to value (LTV) of each individual defaulted file from the long-run
average LTV of the exposure’s pool or grade, or directly through variation of the recovery rate,
e.g. deviation of the realised LGD for a defaulted exposure from the long-run average LGD for the
exposure’s pool or grade. In both cases the deviations observed are partly idiosyncratic but
related also to the current economic conditions. An increase in LTV due to specific investment
strategies (i.e. due to change in risk appetite) of the mortgage owner has a clearly idiosyncratic
nature while an increase in LTV caused by the burst of a house bubble has a clear link with
economic conditions. On the other hand taking the example of shipping finance, a default caused
by the sinking of a ship is clearly idiosyncratic and will generate a lower-than-average recovery,
whereas a default caused by a global drop in maritime freight tariffs is influenced by economic
conditions. In the latter case, in fact, positive economic conditions during the recovery process
may lead to better than average recoveries.
87
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
The calibration of the ELBE estimates to current economic circumstance could be performed by
institutions using different approaches, for example:
•
considering risk drivers in the model that are sensitive to macro-economic and credit factors
relevant for the exposure under consideration. In this way economic conditions will be taken
into account in the application of the ELBE by considering the current value of risk drivers for
the defaulted file under consideration
•
including the relevant macro-economic and credit factors relevant for the exposure under
consideration directly in the model.
These choices have been left to institutions as different approaches may be appropriate to
different circumstances. Anyway the analysis of the relevant economic and credit factors and
their dependence with loss rates should follow the general guidance that will be provided by EBA
in the context of the RTS specifying the nature, severity and duration of an economic downturn
under the mandate set out in Articles 181(3)(a) and 182(4)(a) of the CRR.
Question 7.4: Which approach do you use to reflect current economic circumstances for ELBE
estimation purposes?
7.6.2
Relation of ELBE to specific credit risk adjustments
177.
Where the model used for credit risk adjustments satisfies or can be adjusted to satisfy
the requirements for own-LGD estimates set out in Part Three, Title II, Chapter 3, Section 6 of
Regulation (EU) No 575/2013, institutions may use specific credit risk adjustments as ELBE
estimates.
178.
Where specific credit risk adjustments are assessed individually for a single exposure or a
single obligor, institutions may override the ELBE estimates based on specific credit risk
adjustments, where they are able to prove that this would improve the accuracy of the ELBE
estimates and, moreover, that the specific credit risk adjustments are revised in accordance
with the requirements set in section 6.3 on the calculation of economic loss.
179.
For the purposes of the requirement to duly justify situations where the specific credit
risk adjustments exceed the ELBE estimates in accordance with Article 54(2)(f) of Commission
Delegated Regulation xxx/xxxx [RTS on IRB assessment methodology], institutions should
ensure consistency with the economic loss components described in section 6.3 as well as
with the definition of default set out in Article 178 of Regulation (EU) No 575/2013. In
particular, institutions should take into account, the possible differences in the discounting
rate, the presence of collateral that is not eligible under Article 181 (1)(f) of Regulation (EU)
No 575/2013, different treatments of costs and the application of different definitions of
default.
88
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Explanatory box for consultation purposes
It is proposed that the use of specific credit risk adjustments as ELBE estimates should be limited to
those cases where provisions models meet, or could be adjusted to meet (e.g. by modifying the
discounting rate in use), the prudential requirements for own LGD estimates and the
requirements specified in these Guidelines. An exception to this rule is made allowing institutions
to use individually assessed provisions as a possible reason for override where they are able to
prove that they provide a more accurate estimation than the ELBE estimated by facility grade or
pool. Moreover the evaluation of this accuracy should be made independently of the discounting
rate in use. In fact, the individually assessed provisions should be recalculated in such a way that
all projected cash flows are discounted using the discounting rate prescribed in paragraph 122.
Question 7.5: Do you currently use specific credit risk adjustments as ELBE estimate or as a
possible reason for overriding the ELBE estimates? If so how?
7.7 Specific requirements for LGD in-default estimation
180.
For the purpose of considering the possible adverse change in economic conditions
during the expected length of the recovery processes referred to in Article 54(2)(a) of
Commission Delegated Regulation xxx/xxxx [RTS on IRB assessment methodology] the LGD
should reflect at least downturn conditions, such that the LGD in-default is consistent with the
long-run average LGD for defaulted assets, referred to in paragraph 169, adjusted for
downturn conditions in accordance with section 6.7.
181.
For the purpose of Article 181(1)(h) of Regulation (EU) No 575/2013 the LGD in-default
may need to be increased above the level referred to in paragraph 180 in order to ensure that
the difference between the LGD in-default and the ELBE covers for any increase of loss rate
caused by possible additional unexpected losses during the recovery period.
182.
For the purpose of ensuring that LGD in-default is higher than the ELBE, or is equal to in
limited cases for individual exposures, in accordance with Article 54(2)(d) of Commission
Delegated Regulation xxx/xxxx [RTS on IRB assessment methodology] institutions should
analyse and correct the LGD in-default in those situation where the ELBE obtained using
specific credit risk adjustments, in accordance with paragraph 177, is above the LGD indefault obtained through direct estimation in accordance with Article 54(1)(a) of Commission
Delegated Regulation xxx/xxxx [RTS on IRB assessment methodology].
183.
To the extent that the reasons for overriding the ELBE estimates are relevant also to LGD
in-default a consistent override should be applied also to the LGD in-default in such a way
that the add-on to the ELBE covers for any increase of loss rate caused by possible additional
unexpected losses during the recovery period as in accordance with Article 181(1)(h) of
Regulation (EU) No 575/2013.
89
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
184.
Irrespective of which of the two approaches referred to in point (a) and (b) of Article
54(1)(a) of Commission Delegated Regulation xxx/xxxx [RTS on IRB assessment methodology]
is used for the purposes of estimating LGD in-default institutions should document separately
all of the following:
(a) the break-down of the LGD in-default into its components: the ELBE and the add-on;
(b) the break-down of the add-on into its components:
(i) the downturn conditions component calibrated on the downturn adjustment to the
long-run average LGD as specified in paragraph 180,
(ii) the MoC component, referred to in section 4.4,
(iii) and any component covering for potential additional unexpected losses during the
recovery period referred to in Article 181 (1)(h) of Regulation (EU) No 575/2013.
Explanatory box for consultation purposes
It is proposed that in order to reflect adverse change in economic conditions mentioned in Article
54(2)(a) of the RTS on IRB assessment methodology institutions should reflect at least downturn
conditions in their LGD in-default estimates. This implies that the add-on to the ELBE should
include a component covering for downturn conditions (A component in figure 1) in such a way to
obtain an LGD in-default which is consistent with the long-run average LGD for defaulted assets
adjusted to reflect downturn conditions, i.e. [ELBE + A]= [LRA LGD + C]. In other words, irrespective
of which of the approaches is used for LGD in-default estimation (direct estimation or ELBE + Addon) institutions should first compute the downturn adjustment to the long-run average LGD for
defaulted exposures (C component in figure 1) and then calibrate the add-on downturn
conditions component to the ELBE (A in figure 1) accordingly. The LGD in-default is then obtained
either:
•
directly, increasing the LRA LGD adjusted for downturn conditions (LGD downturn in figure 1,
or LRA LGD + C) by the MoC and additional potential unexpected losses (B component in
figure 1);
•
or as the sum of the ELBE and the add-on, where the add-on is given by summing up the
downturn conditions component (A component in figure 1) to the MoC and the additional
potential unexpected losses if any (B components in figure 1 respectively).
90
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Figure 1: LGD in-default estimation, direct estimation or ELBE + Add-on*
*Where the add-on components to the ELBE include:
A. downturn conditions, where C is the downturn adjustment to the long run average LGD for
defaulted assets and A is the Add-on component to the ELBE to obtain the LGD downturn.
B. MoC + potential additional unexpected losses to the ELBE
91
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
8. Application of risk parameters
8.1 Conservatism in the application of risk parameters
185.
For the purpose of Article 171(2) of Regulation (EU) No 575/2013 institutions should apply
additional conservatism to the outcomes of the process of assignment of exposures to grades
or pools in case of any identified deficiencies related to implementation or application of risk
parameters, especially when those deficiencies relate to data used in the assignment process.
They should do so by establishing a framework that consists of the following phases:
(a) Identification of deficiencies of implementation or application of risk parameters;
(b) Specification of the form of conservatism to be applied and quantification of the
appropriate level of conservatism;
(c) Monitoring of the deficiencies and correcting them;
(d) Documentation.
186.
For the purpose of paragraph 185(a) institutions should have a robust process for
identifying all implementation and application deficiencies in the assignment process,
whereby each deficiency leads to additional conservative treatment in the affected
assignment to a grade or pool. Institutions should consider at least the following triggers for
additional conservatism:
(a) missing data in the current portfolio;
(b) missing updates of financial statements or credit bureau data as referred to in paragraph
66(h);
(c) outdated ratings in the current portfolio; where outdated rating should be understood as
specified in Article 25(2)(b) of Regulation (EU) No xxx/xxxx [RTS on IRB methodology];
(d) missing ratings, whereby an exposure is considered as being within the scope of
application of the IRB model but is not rated by it.
187.
For the purpose of paragraph 185(b) institutions should ensure that the occurrence of any
of the triggers referred to in paragraph 186 results in the adding of additional conservatism to
the risk parameter for the purpose of the calculation of risk weighted assets. Where more
than one triggers occur, the estimate should be more conservative. The additional
conservatism related to each trigger should to the extent possible be proportionate to the
uncertainty in the estimated risk parameter introduced by the trigger.
92
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
188.
Institutions should consider the overall impact of the identified deficiencies and the
resulting conservatism at the level of portfolio covered with the considered model on the
soundness of the assignments to grades or pools and ensure that the own funds requirements
are not distorted due to the necessity for excessive adjustments.
189.
For the purpose of paragraph 185(c) institutions should regularly monitor the
implementation and application deficiencies and the levels of additional conservatism applied
in relation to them. Whenever possible, institutions should take steps to address the
identified deficiencies. Following its assessment, the institution should develop a plan to
rectify the deficiencies within a reasonable timeline, taking into consideration the magnitude
of the impact on the own funds requirements.
190.
For the purpose of paragraph 185(d) institutions should specify adequate manuals and
procedure for applying additional conservatism and should document the process applied in
addressing implementation and application deficiencies. Such documentation should contain
at least the triggers considered and the effects the activation of such a trigger had on the final
assignment to a grade or pool and on the own funds requirements.
Question 8.1: Do you see operational issues with respect to the proposed requirements for
additional conservatism in the application of risk parameter estimates?
8.2 Human judgement in the application of risk parameters
191.
Institutions may use human judgement in the application of the model in the following
cases:
(a) in the application of the qualitative variables in the model;
(b) via overrides of the inputs to the model;
(c) via overrides of the model outputs.
192.
Institutions should specify clear criteria for the use of qualitative model inputs. They
should ensure consistent application of such inputs by all relevant personnel and a consistent
assignment of obligors or facilities posing similar risk to the same grade or pool as required by
Article 171(1)(a) of Regulation (EU) No 575/2013.
193.
For the purpose of Article 172(3) of Regulation (EU) No 575/2013 institutions should
specify the policies and criteria for the use of overrides in the application of the models.
These policies should refer both to possible overrides of inputs and outputs of the models and
should be specified in a conservative manner such that the scale of the conservative override
should not be limited whereas the possibility to decrease the estimate resulting from the
model, either by overriding the inputs or outputs of the model should be limited. In applying
the overrides institutions should take into account all relevant and up-to-date information.
93
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
194.
Institutions should document the scale and rationale of each override. Wherever possible
institutions should specify a predefined list of possible justifications of the overrides to
choose from. Institutions should also store information on the date of override and the
person that performed and approved it.
195.
Institutions should regularly monitor the level and justifications for overrides of inputs
and outputs of the models. They should specify in their policies the maximum acceptable rate
of overrides for each model. Where those maximum levels are breached, adequate measures
should be taken by the institution. The rates of overrides should be specified and monitored
at the level of calibration segment. Where there is a high number of overrides institutions
should adopt adequate measures to improve the model.
196.
Institutions should regularly analyse the performance of exposures in relation to which
an override of input or output of the model has been performed in accordance with Article
172(3) of Regulation (EU) No 575/2013.
197.
Institutions should regularly assess the performance of the model before and after the
overrides of model outputs. Where the assessment concludes that the use of overrides
significantly decreased the model capability to accurately quantify the risk parameters
(‘predictive power of the model’), institutions should adopt adequate measures to ensure the
correct application of overrides.
94
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
9. Re-development, re-estimation and
re-calibration of internal models
198.
Institutions should specify internal policies for re-development, re-estimation, and recalibration of internal models, which consider at least the following potential sources for
triggers:
(a) results of regular review of estimates;
(b) changes in the legal environment;
(c) deficiencies identified by internal audit or the competent authority.
199.
In case material deficiencies are identified by one of the sources, depending on the
severity, a re-calibration, re-estimation or re-development should be triggered and an
appropriate MoC should be applied in accordance with section 4.4. At a minimum, for the
purpose of compliance with Article 43 of Commission Delegated Regulation xxx/xxxx [RTS on
IRB assessment methodology] the institution should describe the applied metrics, thresholds
and accepted deviations for representativeness analysis, discriminatory power analysis,
predictive power analysis and stability analysis.
9.1 Components of regular review of estimates of risk
parameters
200.
For the purpose of performing annual reviews of estimates in accordance with Article
179(1)(c) of Regulation (EU) No 575/2013 institutions should have a framework in place which
includes at least the following elements:
(a) a minimum scope of analyses to be performed, including predefined metrics chosen by
the institution to test model performance and its predictive power;
(b) predefined standards, including predefined thresholds and significance levels for the
relevant metrics;
(c) predefined actions to be taken in case of adverse results in any of the analyses.
201.
For the purpose of paragraph 200(c), institutions should investigate and decide on the
adequate steps in order to remediate identified deficiencies. This may require in particular redevelopment of the model, re-estimation of risk parameters or re-estimation of any model
components.
95
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
202.
The analyses referred to in paragraph 200(a) should at least comprise the following
elements:
(a) a representativeness analysis in order to identify potential differences between the
reference dataset used to estimate the risk parameter and the current portfolio to which
the estimates are applied; this analysis should include the following analyses of any
changes in the portfolio or any structural breaks:
(i) along relevant risk drivers and segmentation drivers used in the rating system;
(ii) due to changes in the underwriting, recovery or default identification process as
well as relevant technical advances;
(iii) due to changes in the scope of application of the model;
(iv) due to structural changes in market and economic conditions.
Where institutions identify significant deficiencies in terms of the representativeness of
the dataset used to estimate risk parameters or where the model’s discriminatory power,
as referred to in point (b), is deteriorating, they should perform the representativeness
analysis as described in the first subparagraph also for the dataset used in model
development.
(b) analysis of the performance of the model and its stability over time; this analysis should:
(i) identify any potential deterioration of the model performance (in particular
discriminatory power) through the comparison of its performance at the time of
the development against its performance on each subsequent observation period
of the extended data set as well as against the predefined thresholds; in
particular should this analysis be performed on relevant subsets, for instance with
and without delinquency days;
(ii) be performed with regard to the whole application portfolio, without any data
adjustments or exclusions; for comparison purposes, the performance at the time
of development must be obtained also for the whole portfolio, prior to any data
adjustments or exclusions;
(iii) be performed according to metrics and standards defined by the institution in
accordance with paragraph 200 and applied consistently over time.
(c) analysis of the predictive power of the model, including at least:
(i) an analysis of whether the inclusion of the most recent data in the dataset used
to estimate risk parameters leads to materially different risk estimates and in
particular:
96
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
 for PD, whether including the most recent data leads to a significant
change in the long-run average default rate; this analysis should take into
account the appropriate redefinition of the period of likely range of
variability of default rates and of the mix of good and bad years, if
necessary;
 for LGD, whether including the most recent data leads to a significant
change in the long-run average LGD or downturn LGD;
(ii) a backtesting analysis, which should include a comparison of the estimates used
for the calculation of own funds requirements against observed outcomes for
each grade or pool.
203.
Institutions should specify conditions when the analyses referred to in paragraph 202
should be performed more frequently than annually. These conditions should include the
specification of events that trigger the analyses such as major changes in the risk profile,
credit policies or relevant IT systems.
204.
For the purpose of performing the tasks referred to in Article 190(2) of Regulation (EU) No
575/2013 institutions should define a regular cycle for full review of the rating systems, taking
into consideration their materiality, covering all aspects in development, estimation of risk
parameters and, where applicable, of model components. This review should include a review
of the selection of the existing and potential risk drivers and assess their significance based on
the predefined standards. The review should also include an assessment of the modelling
approach, its conceptual soundness, fulfilment of modelling assumptions and alternative
approaches. Where the results of this review recommend changes to model design, a
respective re-development of the model should be carried out.
205.
For the purpose of the review specified in paragraphs 200 to 204 institutions should use
consistent rules for data adjustments and exclusions and ensure that any difference in these
processes between the relevant datasets is justified and does not distort the results of the
analysis.
Question 9.1: Do you agree with the proposed principles for the annual review of risk
parameters?
97
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
10. Calculation of IRB shortfall or
excess
206.
Where the calculation for the overall non-defaulted portfolio of Article 159 of Regulation
(EU) No 575/2013 results in an IRB excess, institutions may use this excess to cover for any
IRB shortfall from the overall defaulted portfolio.
Explanatory box for consultation purposes
Article 159 of Regulation (EU) No 575/2013 requires the institutions under the IRB approach to
calculate the difference between expected loss amounts and credit risk adjustments, additional
value adjustments and other own funds reductions for the purpose of own funds recognition.
Article 159 specifies, moreover, that SCRAs on exposures in-default shall not be used to cover
expected loss amounts on other (i.e. non-defaulted) exposures. It has been clarified in Article
73(h) of the RTS on IRB assessment methodology that this in practice means that the amount of
shortfall or excess of provisions should be calculated on an aggregate level for IRB exposures,
separately for defaulted and non-defaulted exposures.
Article 159 specifies that the excess of provisions for defaulted exposures cannot be used to cover
the shortfall of provisions for non-defaulted exposures. However, there is no provision Regulation
(EU) No 575/2013 which prevents the excess of provisions for non-defaulted exposures being
used to cover the shortfall of provisions on defaulted exposures.
207.
For the purposes of adding any IRB excess to Tier 2 in accordance with Article 62 (d) of
Regulation (EU) No 575/2013, where the calculation referred to in Article 159 of Regulation
(EU) No 575/2013 results in an IRB excess for both the defaulted and the non-defaulted
portfolio, the sum of those two IRB excesses should be considered and added to Tier 2 in
accordance with the limit referred to in Article 62(d) of Regulation (EU) No 575/2013.
208.
For the purposes of Article 159 of Regulation (EU) No 575/2013 institutions should not
include partial write-offs in the calculation of general and specific credit risk adjustments.
However, as per Article 166(1) of Regulation (EU) No 575/2013, the calculation of the
expected loss amount for the application of Articles 158 and 159 of Regulation (EU) No
575/2013 should be based on the exposure value gross of value adjustments but net of writeoffs.
Explanatory box for consultation purposes
This has been clarified in the Q&A Question ID 2014_1064. Annex V, Part 2 paragraphs 49 and 50
of Regulation (EU) No 680/2014 (ITS on Supervisory Reporting) clarifies that write-offs are the
amount of principal and past due interest of any debt instrument that an institution is no longer
recognising because they are considered uncollectible, and that they can be caused both by
reductions of the carrying amount of financial assets recognised directly in profit or loss as well as
98
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
by reductions in the amounts of the allowance accounts for credit losses taken against the
carrying amount of financial assets. Such partial write-offs do not constitute impairment,
irrespective of the method (specific loan loss provision or direct reduction of the carrying amount)
chosen to book impairment in the financial statements of the asset, because any amounts
written-back following a derecognition will not impact the carrying amount of the financial asset
(unlike a reversal of impairment losses). For that reasons a partial write off would not be included
in the calculation of general and specifics CRAs.
Question 10.1: Do you agree with the clarifications proposed in the draft GL with regard to the
calculation of IRB shortfall or excess?
Question 11.1: How material would be in your view the impact of the proposed guidelines on
your rating systems? How many of your models do you expect to require material changes that
will have to be approved by the competent authority?
99
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
4. Accompanying documents
4.1 Draft cost-benefit analysis / impact assessment
The EBA considers it adequate to provide an Impact Assessment (IA) which analyses ‘the
potential related costs and benefits’ of the policy provided in the draft guidelines. This analysis
shall provide the reader with an overview of the findings as regards the problem identification,
the options identified to remove the problem and their potential impacts.
The following analysis consists basically of three parts, where the baseline scenario in terms of
current practices and supervisory expectations starts from the analysis performed for the
purpose of the reports on comparability and pro-cyclicality of capital requirements published
by the EBA in December 2013. In terms of the regulatory environment the baseline scenario is
set out by the guidelines specified by CEBS in 2006 (GL 10) under the assumption that this
guidance has been followed by institutions when developing their risk parameter estimation
methodologies. The second part contains the options considered with respect to the major
policy decisions included in the consultation paper. Finally, the draft cost-benefit analysis is
based on the main policy changes in comparison with the currently applicable GL 10. However,
the EBA is aiming to gather more information on the current practices directly from institutions
via a qualitative survey which will be issued during the consultation phase. The objective of this
survey will be the assessment of the scope and severity of potential model changes that will
have to take place in the implementation of these guidelines.
A. Problem identification
The EBA reports on comparability and pro-cyclicality of risk weighted assets (RWA) have
identified significant non-risk based variability of capital requirements calculated in accordance
with the IRB Approach and provided recommendations on regulatory measures that should be
taken in order to achieve greater comparability of risk parameters. All issues that have been
considered while developing these guidelines refer to the identification and/or limitation of
drivers of unjustified RWA variability in the context of PD and LGD estimation and the
treatment of defaulted assets for IRB institutions.
B. Policy objectives
The objective of the guidelines is to establish convergence of institutions’ and supervisory
definitions related to PD and LGD estimation and the treatment of defaulted assets as well as the
convergence of institutions’ methodological choices in developing PD and LGD models where
these choices are considered to be drivers of unjustified RWA variability. The guidelines are
complementary to the RTS on assessment methodology (EBA/RTS/2016/03) where some
technical choices related to PD and LGD estimation as well as to the treatment of defaulted assets
100
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
are already regulated. In particular, provisions related to data requirements, the estimation of the
long-run average PD and the calculation of the long-run average LGD as simple average with
respect to the number of defaults refer to the RTS on assessment methodology as a starting
point.
The guidelines aim to define common notions and criteria in the major policy fields including:
a) A framework for the margin of conservatism
b) PD: One-year-default rates (frequency of motoring)
c) PD: Long-run average default rates
d) LGD: Additional drawings after default and interest and fees capitalised after the
moment of default
e) LGD: Discount rate
f)
LGD: Incomplete recovery processes
C. Baseline scenario
The baseline scenario is specified on the basis of data collected from competent authorities in
2013 for the purpose of the report on comparability of supervisory practices that are summarised
in the table below.
Paragraphs Subject
in the CP
PD estimation
23-35
Margin of
conservatism (i.e.
rules concerning the
conservatism in the
PD parameter and
rating systems for
data or model
weaknesses)
Findings from Report on the comparability of
supervisory rules and practices
In general, most CAs confirm the requirement of that
conservatism should be applied for data and model issues
and none of the CAs provides guidelines on the level of
conservatism that is expected. As such, the adequacy of
the level of conservatism is mainly assessed on a case-bycase basis and less than 30% of CAs has rules on this
topic, most being non-public. These rules concern the
following aspects: (i) one CA requires that a margin of
conservatism is applied for low default portfolio models
and for models primarily based on human judgement, (ii)
one CA requires an explicit statistical approach to assess
the margin of conservatism combined with a qualitative
adjustment when the defaults are fewer than 20, (iii) one
CA requires that conservatism is applied to address the
‘seasonal peak’ for mortgages, (iv) another one mentions
that a non-compliant definition of default should be
addressed by a margin of conservatism, (v) another
mentions that institutions should have a methodology to
assess the margin taking into account the results of the
validation and the results of self-assessment; and (vi)
finally one CA mentions that a supervisory add-on will
compensate for institutions’ lack of conservatism.
101
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
21, 45 and
48-63
Data requirements
(i.e. rules regarding
the use of internal
default rates,
mapping with external
ratings or use of
statistical PD model)
and calculation of
observed default rates
59-63
Historical observation
period (i.e. rules on
the business
cycle/’downturn
period’)
64-67
Risk drivers and rating
criteria (i.e. rules
concerning the
assessment of the
choice of variables
that influence the PD
estimation)
75-79
Design of grades or
pools
One-third of the CAs specifies rules concerning the PD
estimation approach, almost all non-public. In general,
the use of internal default rates or statistical approaches
is promoted or requested as long as the data are relevant
and representative. For low default portfolios, the
mapping with external rating or expert models is usually
tolerated, but (i) with the use of an additional level of
conservatism, or (ii) where benchmarking should be
performed with other approaches, and where again the
institutions are encouraged to collect data in order to
develop statistical models in due time. Some CAs
however allow for all approaches without restriction.
Whereas the majority of CAs reported rules on the topic,
these rules were public and binding in only two cases.
However these rules are usually not specific, but rather
general and ‘principle based’. A wide variety in
requirements has been observed; examples are that (i)
institutions should generally include good and bad
economic periods, or (ii) periods when higher credit
losses are experienced, or (iii) should include at least one
recession period, or (iv) should cover a complete
economic cycle (good and bad years), or (v) institutions
are required to include specific periods in the datasets
(e.g. 1991–2008). Only few CAs specify the years of
reference for a recession period or a complete cycle.
Some CAs mention that different weights can be applied
to data (higher weights to more recent data) if
adequately motivated.
Approximately one-third of CAs have rules on this topic,
most being non-public and general, in the sense that the
most relevant variables should be included in the model
and missing variables are challenged. In two Member
States there is a requirement that there should be risk
drivers or rating criteria on some key groups of variables
(such as characteristics of the borrower, terms of the
transaction, collateral, unpaid amounts, etc). When it
comes to assessing the information value of the risk
drivers, some Member States specify a threshold for the
p-value of the regression, whereas others specify more
general rules regarding the adequacy of the
discriminatory power.
More than 40% of CAs has rules concerning the number
of grades, which are usually public and binding. The
majority sticks to the minimum number of grades fixed by
the CRD (seven plus default); one CA increases the
number to 10 for wholesale portfolios.
Related rules require in some Member States that the
number of pools and grades should be high enough to
allow adequate quantification and validation. In one
102
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
75-79
Rating philosophy (i.e.
rules specifying how
stable rating
assignments should
be during the
economic cycle or
how much they
should change with
the economic context
(Point-In-Time (PIT),
Through-The-Cycle
(TTC) or hybrid
philosophies)
80
Calibration sample
(i.e. rules concerning
the data used for
calibration of the PD
estimate)
LGD estimation
89-92
General requirements
with regard to
estimation
methodologies
Member States, the concentration across the rating scale
should be assessed by the Hirschman-Herfindahl index. In
another Member State, the concentration in one grade
should not be greater than 30% unless the grade cover a
reasonably narrow PD band. Another CA applies a 25%
limit for the wholesale portfolios. These rules however do
not apply to low default portfolios, where a case-by-case
approach is typically allowed.
Only very few CAs appeared to have rules on this topic,
mostly non-public, and only one CA publicly requires
ratings to be TTC. One CA explicitly assesses the stability
of the ratings and one other specifically allows hybrid
systems. Some of the other CAs note that they implicitly
allow for all rating philosophies, leaving the choice to
institutions. The CAs seem to focus on the inclusion of all
relevant and recent information regarding the credit
quality of obligors.
The CAs have also been asked whether they have rules
concerning the dynamics of the transitions of exposures
or clients among rating classes. However, only three CAs
reported rules on the topic, all non-public. This means in
particular that they monitor/challenge the migration
matrices in order to assess their stability, at different
stages of the economic cycle. For two CAs, analysis
regarding the stability of migration matrices is only
conducted during the approval process of the IRB
Approach.
Around one-third of the Member States have rules on this
topic, some being public and binding and some nonpublic. These rules usually mention the requirement of
5 years of representative data. When a shorter period is
used, a margin of conservatism must usually be added to
address the data issue. The answers however appeared
incomplete for several CAs and therefore no robust
conclusions could be drawn.
The majority of Member States do not have any specific
rules with regard to estimation methodologies apart from
those specified in GL 10 but it was indicated by several
respondents that mostly (of even only) workout LGD is in
use. Only in 2 jurisdictions there were public and binding
rules in that regard, where workout LGD was prescribed
in one of them for mortgage portfolios and in the other
workout LGD was expected for corporate and retail
portfolios while market LGD was allowed for large
corporate, institutions and sovereign exposures. A few
Member States apply certain rules with regard to cure
rates: one of them explicitly requires institutions to
identify cure rates and another forms such requirement
103
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
112-127
Calculation of
economic loss for the
purpose of workout
LGD, including
discounting factor
134-138
Length of workout
period and treatment
of incomplete
workout cases
140-144
Risk differentiation
128-133
Length of observation
period
where there is potentially higher rate of technical
defaults. In any case the definition of cure is left for the
institutions to be specified.
Only several Members States have any specific rules with
regard to the calculation of economic loss and realized
LGD and most of them are not public and cover only
selected aspects such as for instance conservative
approach to cost calculation and allocation. One of the
CAs defined minimum types of data to be collected by the
institution for the purpose of calculation of realized LGD
while another specified that in cases of collateral
repossession, institutions have to estimate haircuts on
the value of the collateral considering the potential sale
value and the time to sale.
4 CAs specified specific rules for the discounting factor
which include the following approaches: (i) the discount
rate that is applied shall reflect the uncertainty at the
time of default; (ii) discount rate should not be lower
than risk free rate; (iii) a risk premium calculated using an
internal model or, in its absence, 400 basis points over
the base rate; (iv) minimum expectation of a 9% discount
rate. Other CAs base on the principles specified in GL 10
or in BCBS guidelines.
A few CAs specified some general rules with regard to the
length of workout period, whereas one CA considers 95%
of recovery rate to be a specific point to terminate
workout process for recovery curve calculation purpose.
In addition, one CA requires institutions to take into
account at least 2 complete recovery cycles. With regard
to the treatment of incomplete cases most of the
Member States do not have any specific rules. Those few
that do have such rules require including such cases in the
calculation and allow estimation of future cash flows.
Only 2 CAs have specific rules in terms of the granularity
of risk differentiation and in both cases they are not
public. Institutions are required to ensure adequate
segmentation within each portfolio in order to ensure a
proper risk differentiation. More rules exist with regard to
explanatory variables, in 3 jurisdictions such rules are
public and binding whereas in 6 other they are not public
but part of supervisory practices. While in some cases
the rules refer to certain risk drivers that have to be used
in model development, other CAs require that banks use
the most relevant variables and that they adequately
reflect the recovery processes.
While rules regarding the length of the observation
period exist in several jurisdictions, in a few of those
cases they are only reflective of the minimum periods
specified in the CRR. Other requirements specified by
104
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
145-155
Collaterals and
guarantees
156
Downturn LGD
23-35
Margin of
conservatism
individual Member States include the following: (i)
covering a complete business cycle; (ii) covering at least 2
complete recovery cycles. In a few cases different
weighting is allowed, for instance to address structural
changes in data, but such weighting has to be sufficiently
justified and conservatism has to be ensured. In addition,
one CAs required including downturn period of 1990s.
The rules on the treatment of collaterals and guarantees
exist in several jurisdictions but only in half of them these
rules are public. The examples of such rules include the
following: (ii) estimates have to be made for
homogeneous pools of collateral, where similar
recoveries can be expected and shown; (ii) FX risk must
be considered; (iii) banks must establish internal
standards for collateral management consistent with
those required under the standardized approach. With
regard to valuation there are various requirements across
jurisdictions which range from requiring that the
valuation of real estate collateral is done by an
independent appraiser to allowing the use of statistical
methods in determining the value of the real estate. In
some cases it is required to apply specific haircuts which
are specified differently by different Member States. In
addition, a few CAs specified rules with regard to
minimum frequency of revaluations which again differ
across jurisdictions. While one CA requires annual reviews
for real estate collaterals and quarterly for financial
collaterals, other allow less frequent reevaluations, at
least in some cases.
In general, the calibration of the LGD parameter is
affected by the downturn LGD calibration. However, only
35% (7 CAs) of the CAs define any rules concerning the
methodology of downturn LGD. Among those CAs in only
one case the rule is public and binding. Moreover, most of
the CAs confirm that banks should base their downturn
LGD estimates on historical scenarios and moreover check
for conservativeness of the estimation made at the
institution level.
While a few CAs apply some rules in terms of the margin
of conservatism they are usually not very prescriptive in
terms of the quantification of MoC and are mostly
focused on the general types of weaknesses that the MoC
should address (data, methodologies); one CA requires
specifically that MoC should cover additionally significant
differences between the debtors and their guarantors
positive correlations and currency mismatches between
exposures and collaterals. In one Member State it is
required that MoC should be applied on top of the
estimates and another specifies that can be established
105
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
as an LGD floor (e.g. 45% in the case of large corporate)
or specific add-ons (e.g. stressing the probability of
incompletes).
Defaulted exposures
157-179
ELBE
157-171
and 180184
LGD in-default
Only 35% of the CAs (7 CAs) define a rule concerning the
methodology of the best estimate of expected losses on
defaulted assets. Moreover, the rules specified show
divergence of practices, for example, some refer to the
GL10 rules, others align the methodology to the LGD for
non-defaulted exposures and some simply use provisions.
Among those CAs which specify a rule only 3 confirm to
have this rule public and, in particular, in two cases the
rule is also binding.
Only 25% of the CAs (5 CAs) define a rule concerning the
methodology of LGD in-default calculation. Among those
CAs in only one case this rule is public and binding. In
terms of approach used in most of the cases LGD indefault is obtained as ELBE + add-on or as LGD downturn.
D. Options considered
This section presents the assessment of the technical options considered in the CP. Under each
option, the potential advantages and disadvantages of the options together with potential costs
and benefits are discussed.
Framework for margin of conservatism
Current regulatory framework does not provide detailed criteria related to data and method
deficiencies for which institutions have to provide margin of conservatism according to Article
179 of the CRR. Moreover, there are currently no provisions regarding the level at which MoC
should be identified and there is no guidance on the quantification of MoC for certain triggers
or in general.
Regarding the general framework for MoC the following options were considered:
(a) non-exhaustive long list of triggers for MoC as part of the guidelines (including
recommendations for the according appropriate adjustments):
 Pros: provides a more harmonised approach towards the triggers that require MoC
and the method to quantify the impact;
 Cons: could lead to less suitable approaches for individual models.
(b) MoC categories with minimum list of triggers in the guidelines:
 Pros: provides a base for a more harmonised reporting on the level of margin of
conservatism, but leaves room for distinct approaches at the same time;
 Cons: different approaches could still lead to different levels of MoC.
106
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
It has been decided in favor of option (b).
Regarding the level at which MoC should be identified the following options were discussed:
(a) quantification and application of MoC at the level of calculation of the long-run average
default rate:
 Pros: simple approach;
 Cons: can lead to significantly different capital requirements where direct estimates
of risk parameters for individual obligors or exposures on a continuous scale,
calibrated to a central tendency, are used.
(b) quantification and application of MoC at the level it is identified but reporting with respect to
the final estimate of risk parameter used for own funds requirements:
 Pros: ensures a harmonised approach and common notion of long-run average default
rate excluding MoC.
It is a common practice to add MoC to the long run average default rate or the central tendency
(which is quite often specified based on the long run average default rate of the considered
portfolio). However, Article 179 of the CRR says that an institution shall add to its estimates a
margin of conservatism, which is literally not done if adding it to the long run average default
rates (or central tendency) and using these as input to a calibration method which might be nonlinear. In the light of continuous PD estimates where there might be only 1 obligor per grade,
developing a confidence interval (to specify the margin of conservatism) might not be straight
forward and also to prevent potential breaches of monotony of the estimation of risk parameters
used for own funds calculation it has been decided to introduce option (b) in the guidelines.
One-year-default rates (frequency of motoring)
Regarding the prescribed frequency of monitoring one year default rates in paragraph 53 the
following policy options have been considered:
(a) the one-year default rate calculated at least at a monthly frequency for all retail exposures,
and at least quarterly for all other exposures:
 Pros: ensures up-to date information for the purpose of internal risk managements
and allows identification of changes in the trends in a timely manner;
 Cons: for low default portfolios probably no new information contained;
 Cons: could be overly burdensome.
(b) the one-year default rate calculated at least quarterly for all exposures:
 Pros: less burdensome but at the same time ensures minimum frequency of
monitoring that allows identification of any seasonal effects;
 Cons: obligors that default in less than three months after credit origination could
be omitted in the default rate.
107
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
(c) the one-year default rate calculated at least at a quarterly frequency for all retail exposures,
and at least at semi-annually for all other exposures:
 Cons: obligors that default in less than half a year (for non-retail) after credit
origination could be omitted in the default rate.
Option (b) has been chosen as the option which balances the burden on the institutions and
ensures a base for comparability of analysis and reporting of default rates.
Long-run average default rates
Article 180(1)(a) of the CRR requires institutions to estimate PDs by obligor grade from long-run
averages of one-year default rates. Therefore the notion of the long-run average default rate is
crucial and a number of options have been considered. The main policy options discussed can
be summarized as follows:
(a) Long-run average default rate estimated as the observed average default rate multiplied by
an adjustment-factor where appropriate and subject to certain triggers and limitations:
 Pro: this approach would provide a high level of harmonization;
 Cons: might be overly complex;
 Cons: difficult to implement for both institutions and competent authorities.
(b) Long run average default rate estimated starting from the observed average default rate
and adjusted subject to the criteria listed in paragraph 61 (a)-(c);
 Pros: leaves flexibility where appropriate to estimate the likely range of variability of
one-year default rates;
 Cons: may lead to excessive flexibility and insufficient harmonisation.
(c) Long run average default rate estimated as the observed average default rate and additional
MoC where the according historical observation period does not contain a downturn period.
 Pro: simple;
 Cons: could lead to pro-cyclicality in capital requirements where downturn years are
overly represented in the historical observation period.
Option (a) required the estimation of a long-run adjustment factor, which would be a particular
kind of appropriate adjustment, for which also a MoC should be applied (consistent with the
approach proposed in the chapter on margin of conservatism). Under this option the
appropriate adjustment should lead to an “expected” value of a one year PD, where MoC
should account for the additional uncertainty due to the data or method deficiencies. The
bearing point in the discussion related both to options (a) and (b) was whether the estimated
long run average default rate can be lower than the observed average default rate. Under
option (a) such result would have been allowed subject to very restrictive analysis and only
following certain circumstances (e.g. historical data only from DT-years but improvement of
economic conditions can be proven over the last 3 years).
108
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
Option (c) has the advantage of being the simplest one but has been criticized as having the
potential of leading to pro-cyclicality. Eventually, option (b) was chosen as a compromise. It
should be noted that possible benchmarks or floors have been discussed under all options but
were softened again due to concerns regarding pro-cyclicality.
Additional drawings after default and interest and fees capitalised after the moment of default
As outlined in the explanatory box two approaches have been considered for the treatment of
additional drawings after default and interest and fees capitalized after the moment of default.
(a) Interest and fees capitalised after the moment of default as well as the according cash flows
not to be taken into account in the determination of economic loss
 Pros: ensures that the reported EAD at the date of default equals the EAD for the
purpose of calculation of realized LGD;
 Pros: simple approach with less burden of data collection;
 Cons: may not be sufficiently prudent as the recoveries related to the interest and fees
calculated after default would decrease the economic loss;
 Cons: due to differences between the discounting rate and the interest rate applicable
after default the value of money in time and the recoveries related the amount
outstanding at the moment of default would not be reflected correctly.
(b) Interest and fees capitalised after the moment of default as well as the according cash flows
to be taken into account in the determination of economic loss.
 Pros: ensures accurate and consistent calculation of realized LGD.
 Cons: requires a broader scope of data collection.
Discounting rate
The discounting rate has been recognised as one of the major drivers of the lack of RWA
comparability across institutions and for this reason it is proposed that these guidelines should
provide detailed guidance on this aspects. The broad variety of practices requires therefore clear
guidance on what should be reflected by the discounting factor and how it should be applied. In
this regard the following options have been considered:
(a): Original effective interest rate: discounting factor is derived from facility specific interest
rates:
 Pros: consistency with international accounting standards;
 Cons: lack of comparability of losses within and across institutions as the interest may
depend on institutions pricing strategies.
(b) Risk free rate + add-on: the add-on reflects the risk premium for the uncertainty related to the
recoveries:
109
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
 Pros: ensures simplicity and comparability and independence from the own credit
standing of the institution;
 Pros: consistency with GL10 and BCBS guidance;
 Cons: if the risk premium is estimated by institutions the harmonization may still not
be sufficient.
(c) Funding costs + add-on: discounting factor reflects the funding costs of the institution and an
appropriate risk premium reflecting the uncertainties associated with the receipt of recoveries
with respect to a defaulted exposure:
 Pros: consistency with GL10 and BCBS guidance where it is proposed that discount
factor should capture the opportunity costs of holding the defaulted asset during the
workout period; better reflects market conditions;
 Cons: lack of comparability could stem from different spread computation
methodologies;
 Cons: institutions from the Member States with worse credit standing may be
systematically penalised.
The application of option (a) would imply that different discounting factors would apply to
different products or even individual exposures. These differences may be very significant even
within one institution and would reflect not only the risk of specific clients or transactions but also
specific marketing policies as well as market and economic conditions at the moment of
origination of a loan. However, if original effective interest rate was used it would remain
unchanged throughout the lifetime of a specific exposure.
Option (b) was chosen as the preferred option. In addition, in order to ensure a harmonised
approach it has been specified that the risk free rate should be chosen as the 1-year EURIBOR or a
comparable interbank rate in the countries from outside the Eurozone (relevant X-IBOR rate) at
the moment of default and the add-on should be calibrated to 5%.
Incomplete recovery processes
The treatment of incomplete recovery processes for the purpose of calculating the average
realized LGD as referred to in Article 181 (a) of the CRR the following options have been
considered:
(a) Institutions may estimate future recoveries for the periods further than the maximum length
of the recovery process only if these recoveries will stem from the realisation of the existing
collaterals:
 Pros: more accurate in some cases, especially where individual case-by-case
assessment is applied and there is high probability that the recovery on a given
exposure will be realised;
 Cons: often not enough data to estimate recoveries for further periods;
110
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
 Cons: less strict approach that allows more subjective and less comparable estimates
and may lead to disregarding in practice the effect of the maximum length of the
recovery process.
(b) Institutions should not estimate any future recoveries for the periods further than the
maximum length of the recovery process:
 Pros: addresses the problem that institutions cannot present reliable estimates for
further periods due to insufficient data;
 Pros: where the collateral has not been realised within the specified period it may
indicate some problems with the collateral that could prevent its realisation;
 Cons: less flexible and hence in some cases less accurate, as it is not possible to include
future expected cash flows even if there is high probability they will be realised (but if
the maximum period is defined appropriately this inaccuracy should not be
significant);
 Cons: Cash flows from collaterals, if they exist, are usually more significant than other
cash flows and at advance stages of recovery processes can be predicted on an
individual basis.
Option 1 (b) was chosen as the most appropriate.
E. Cost-Benefit Analysis
The guidance given in these guidelines regarding the development and maintenance of risk
parameter estimation methodologies as well as regarding the treatment of defaulted assets
will affect all areas of modelling PD, LGD, ELBE and LGD-in default. Therefore it is expected that
these GL will lead to model changes. However detailed assessment of the costs for institutions
of these model changes and their impact on capital requirements is not possible as the current
flexibility of the IRB Approach does not allow a definition of a common baseline scenario
regarding definitions and current modelling choices from a institutions perspective. It is
expected that the impact of these GL on individual institutions will vary depending on the
currently implemented solutions.
However, the main costs of implementation of these draft GL are considered to have the
nature of one-off costs covering:
•
the training of the staff on these draft GL,
•
the redevelopment of the models,
•
the IT specification and implementation of the reviewed models,
•
the set-up of monitoring reports where the drat GL contain additional monitoring
requirements,
•
the costs for the regulatory approval process.
111
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
As the latter type of costs will depend on the severity of the expected model changes, the EBA
plans to issue a qualitative survey to institutions to assess the amount and severity of model
changes expected. However, when analyzing these costs of implementation it has to be kept in
mind that the other regulatory products, in particular the RTS on assessment methodology and
the RTS on materiality threshold, within EBAs review of the IRB Approach will also trigger
material model changes, which are expected to be handled together with the model changes
arising from these draft GL to the extent possible.
Before collecting more detailed data on the current practices of the institutions the expected
impact of these GL can be assessed on the basis of the changes proposed with regard to the
currently applicable GL 10, which can serve as a proxy to assess the nature of the expected
changes. The following analysis outlines the major policy changes with respect to the GL 10 and
provides initial indication of the direction of the severity of model changes expected.
The impact of the GL may be assessed by analyzing the scope of the changes in comparison to the
currently applicable Guidelines on the implementation, validation and assessment of Advanced
Measurement (AMA) and Internal Ratings Based (IRB) Approaches (so called GL 10) published by
CEBS in April 2006. It is planned that relevant parts of GL 10 will be repealed when the new EBA
GL come into force. Hence the changes in the policy reflected in the EBA GL relatively to GL 10 will
indicate the scope of the changes that will have to be introduced in the rating systems of the
institutions.
In principle GL 10 were more general than the currently proposed draft GL and provided more
description of the possible approaches and challenges related to them rather than strict,
normative rules. The currently proposed GL are more specific in most of the addressed areas and
therefore even where there no explicit change of policy is proposed some changes may still be
necessary in order to comply with the more detailed requirements.
The main areas where an explicit change in policy relatively to GL 10 is proposed include the
following:
a. Discounting factor – the draft GL is based on the same concept as GL 10 that discounting
rate should include risk-free rate and risk premium but propose a fixed value of the risk
premium and there is no possibility for institutions to estimate it themselves for specific
portfolios. The GL are less flexible with regard to haw the uncertainty should be addressed
and suggest that it is no longer possible to include this uncertainty elsewhere, for instance
in the recovery cash flows, instead of the discounting factor.
Currently various approaches are used by institutions in practice for determining the
discounting rate ranging from the use of discounting factors based on effective interest
rates of the underlying loans, different add-ons with wide range of values on top of
different underlying internal and external interest rate benchmarks to fixed discounting
rates reflecting downturn conditions. In the light of these differences it is expected that
the policy proposal included in the draft GL may have significant impact on the calculation
of the realised LGDs at least for some of the institutions.
112
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
b. Recovery processes with positive outcome – GL 10 recognise that some defaulted
positions may generate no loss, or may even have positive outcomes in the recovery
process and specify that the estimated LGD used to calculate capital requirements must
not be less than zero. However GL 10 do not prescribe any specific treatment of individual
positive realized LGDs other than that these cases should be monitored. The proposed
draft GL introduce the rule that realized LGDs should be floored to zero at individual level.
It is expected that this policy should not have significant impact on the LGD estimates in
general as these estimates have to reflect downturn conditions. However, for some
specific portfolios that may systematically lead to positive outcomes, such as some leasing
portfolios, this proposal may lead to significant change of the model.
c. LGD estimation methodologies – GL 10 specify that the Market LGD and Implied Market
LGD which are based on market data on the prices of certain instruments are possible
methodologies for estimating LGD in some cases, especially where internal data is not
sufficient and capital markets are deep and liquid. However, it was also recognised already
in GL 10 that these methods could potentially be used only in specific circumstances and
that LGD estimates based on an institution’s own loss and recovery experience (so called
workout LGD) should in principle be superior to other types of estimates. According to the
draft GL methodologies based purely on market data will no longer be allowed as these
methodologies do not meet certain CRR requirements, in particular those related to the
specification of the observation period and representativeness of data. Market data could
only be use to supplement internal data that reflects own experience of the institution.
In the light of the most recent proposals of the Basel Committee for Banking Supervision
(BCBS) with regard to low default portfolios and in particular the proposals for limiting the
scope of application of the IRB approaches it is expected that the impact of the proposed
GL by the date of its application should not be significant. It seems that market based
methodologies are not widely used across the EU, however, the full assessment of the
impact of the policy will only be possible when more data will be gathered on this aspect.
d. LGD in-default – GL 10 seem to allow that downturn adjustment may not be part of LGD
estimation for defaulted exposures as it states that downturn conditions should be taken
into account in measuring the possibility of additional unexpected losses during the workout period if they are relevant to a certain type of exposures. The currently proposed draft
GL are clear that LGD in-default should comply with all requirements for LGD estimation
and therefore it should also reflect downturn conditions.
As the currently proposed GL provide more prescriptive requirements with regard to
estimating LGD for defaulted exposures it may lead to necessity to adjust the calibration of
some of the models.
The above analysis does not include the analysis of the requirements for the estimation of
downturn LGD which may potentially be another aspect where the change in the policy may be
proposed. These aspects however are not currently covered by the proposed text of the GL. The
EBA continues working on the aspects related to downturn adjustment by considering together
the RTS on the nature, severity and duration of economic downturn and the possible additional
113
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
section of the GL that will clarify how to apply the requirements of the RTS to derive the
downturn LGD. The respective proposals will be presented and consulted at the later point in
time.
114
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
4.2 Overview of questions for consultation
4.1: Do you agree with the proposed requirement with regard to the application of appropriate
adjustments and margin of conservatism? Do you have any operational concern with respect to
the proposed categorization?
5.1: Do you see any operational limitations with respect to the monitoring requirement proposed
in paragraph 53?
5.2: Do you agree with the proposed policy for calculating observed average default rates? How
do you treat short term contracts in this regard?
5.3: Are the requirements on determining the relevant historical observation periods sufficiently
clear? Which adjustments (downward or upward), and due to which reasons, are currently
applied to the average of observed default rates in order to estimate the long-run average default
rate? If possible, please order those adjustments by materiality in terms of RWA.
5.4: How do you take economic conditions into account in the design of your rating systems, in
particular in terms of:
d. definition of risk drivers,
e. definition of the number of grades
f. definition of the long-run average of default rates?
Question 5.5: Do you have processes in place to monitor the rating philosophy over time? If yes,
please describe them.
Question 5.6: Do you have different rating philosophy approaches to different types of
exposures? If yes, please describe them.
Question 5.7: Would you expect that benchmarks for number of pools and grades and maximum
PD levels (e.g. for exposures that are not sensitive to the economic cycle) could reduce unjustified
variability?
6.1: Do you agree with the proposed principles for the assessment of the representativeness of
data?
6.2: Do you agree with the proposed treatment of additional drawings after default and interest
and fees capitalised after the moment of default in the calculation of realised LGDs?
6.3: Do you agree with the proposed specification of discounting rate? Do you agree with the
proposed level of the add-on over risk-free rate? Do you think that the value of the add-on could
be differentiated by predefined categories? If so, which categories would you suggest?
6.4: Do you agree with the proposed approach with regard to the specification of historical
observation period for LGD estimation?
6.5: Do you agree with the proposed treatment of incomplete recovery processes in obtaining the
long-run average LGD?
6.6: Do you agree with the proposed principles on the treatment of collaterals in the LGD
estimation?
115
CP ON GLS ON PD ESTIMATION, LGD ESTIMATION AND TREATMENT DEFAULTED ASSETS
6.7: Do you agree with the proposed treatment of repossessions of collaterals? Do you think that
the value of recovery should be updated in the RDS after the final sale of the repossessed
collateral?
6.8: Do you think that additional guidance is necessary with regard to specification of the
downturn adjustment? If yes, what would be your proposed approach?
7.1: Do you agree with the proposed approach to the ELBE and LGD in-default specification? Do
you have any operational concerns with respect to these requirements? Do you think there are
any further specificities of ELBE and LGD in-default that are not covered in this chapter?
7.2: Do you agree with the proposed reference date definition? Do you currently use the
reference date approach in your ELBE and LGD in-default estimation?
7.3: Do you agree with the proposed approach with regard to the treatment of incomplete
recovery processes for the purpose of estimating LGD in-default and ELBE?
7.4: Which approach do you use to reflect current economic circumstances for ELBE estimation
purposes?
7.5: Do you currently use specific credit risk adjustments as ELBE estimate or as a possible reason
for overriding the ELBE estimates? If so how?
8.1: Do you see operational issues with respect to the proposed requirements for additional
conservatism in the application of risk parameter estimates?
9.1: Do you agree with the proposed principles for the annual review of risk parameters?
10.1: Do you agree with the clarifications proposed in the guidelines with regard to the calculation
of IRB shortfall or excess?
11.1: How material would be in your view the impact of the proposed guidelines on your rating
systems? How many of your models do you expect to require material changes that will have to
be approved by the competent authority?
116